oops, forgot to document the /search command and had some lint issues:
- removed unused imports
- fixed template literal type issues
- documented /search in slash-commands.md
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
been working on this for a while now. basically it lets you search messages
across all your connected platforms (whatsapp, telegram, discord etc) from
one place.
the main idea is that same person on different platforms gets linked together
so when you search for "what did sarah say about the deadline" you get results
from everywhere she messaged you.
whats included:
- new contacts module with sqlite storage for the unified graph
- auto-linking by phone number (high confidence) and name similarity
- /search command in chat + clawdbot search CLI
- clawdbot contacts CLI for managing the graph manually
all tests passing (77 tests)
This fixes the msteams probe which otherwise incorrectly assumes teams is not working.
The @microsoft/agents-hosting SDK's MsalTokenProvider automatically appends /.default to all scope strings in its token acquisition methods (acquireAccessTokenViaSecret, acquireAccessTokenViaFIC, acquireAccessTokenViaWID, acquireTokenWithCertificate in msalTokenProvider.ts). This is consistent SDK behavior, not a recent change.
The current code is including .default in scope URLs, resulting in invalid double suffixes like https://graph.microsoft.com/.default/.default. I am not sure how the .default postfixed worked in the past for you if I am honest.
This was confirmed to cause Graph API authentication errors. Removing the .default suffix from our scope strings allows the SDK to append it correctly, resolving the issue. I confirmed it manually on my teams setup
Before: we pass .default -> SDK appends -> double .default (broken)
After: we pass base URL -> SDK appends -> single .default (works)
Co-authored-by: Christof Salis <c.salis@vertifymed.com>
Based on actual Flawd deployment experience:
- Proper fly.toml configuration with all required settings
- Step-by-step guide following exe.dev doc format
- Troubleshooting section with common issues and fixes
- Config file creation via SSH
- Cost estimates
In containers, PIDs can be recycled quickly after restarts. When a container
restarts, a different process might get the same PID as the previous gateway,
causing the lock check to incorrectly think the old gateway is still running.
This fix adds isGatewayProcess() which verifies on Linux that the PID actually
belongs to a clawdbot gateway by checking /proc/PID/cmdline. If the cmdline
doesn't contain 'clawdbot' or 'gateway', we assume the lock is stale.
Fixes gateway boot-loop in Docker/Fly.io deployments.
- Add fly.toml configuration for Fly.io deployment
- Add docs/platforms/fly.md with deployment guide
- Uses London (lhr) region by default
- Includes persistent volume for data storage
- Add latency metrics to summarizeText and textToSpeech functions
- Add /tts_status command showing config and last attempt result
- Add /tts_summary command for feature flag control
- Fix atomic write to clean up temp file on rename failure
- Add timer.unref() to prevent blocking process shutdown
- Add unit tests for validation functions (13 tests)
- Update README with new commands and features
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Add /tts_limit command to configure max text length (default 1500)
- Auto-summarize long texts with gpt-4o-mini before TTS conversion
- Add truncation safeguard if summary exceeds hard limit
- Validate targetLength parameter (100-10000)
- Use conservative max_tokens for multilingual text
- Add prompt injection defense with XML delimiters
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Integrate message_sending hook into Telegram delivery path
- Send text first, then audio as voice message after
- Add /tts_provider command to switch between OpenAI and ElevenLabs
- Implement automatic fallback when primary provider fails
- Use gpt-4o-mini-tts as default OpenAI model
- Add hook integration to route-reply.ts for other channels
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Remove sag CLI dependency
- Add direct ElevenLabs API integration via fetch
- Add OpenAI TTS as alternative provider
- Support multi-provider configuration
- Add tts.providers RPC method
- Update config schema with OpenAI options
- Bump version to 0.2.0
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Add a new extension that provides automatic text-to-speech for chat
responses using ElevenLabs API.
Features:
- `speak` tool for converting text to voice messages
- RPC methods: tts.status, tts.enable, tts.disable, tts.convert
- User preferences file for persistent TTS state
- Configurable voice ID, model, and max text length
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Address reviewer feedback: slash commands now use the same
hasExplicitConfig check as regular messages, so unlisted
channels are allowed under groupPolicy: "open" for both
message handling and slash commands.
## Summary
Fix Slack `groupPolicy: "open"` to allow unlisted channels even when `channels.slack.channels` contains custom entries.
## Problem
When `groupPolicy` is set to `"open"`, the bot should respond in **any channel** it's invited to. However, if `channels.slack.channels` contains *any* entries—even just one channel with a custom system prompt—the open policy is ignored. Only explicitly listed channels receive responses; all others get an ephemeral "This channel is not allowed" error.
### Example config
```json
{
"channels": {
"slack": {
"groupPolicy": "open",
"channels": {
"C0123456789": { "systemPrompt": "Custom prompt for this channel" }
}
}
}
}
```
With this config, the bot only responds in `C0123456789`. Messages in any other channel are blocked—even though the policy is `"open"`.
## Root Cause
In `src/slack/monitor/context.ts`, `isChannelAllowed()` has two sequential checks:
1. `isSlackChannelAllowedByPolicy()` — correctly returns `true` for open policy
2. A secondary `!channelAllowed` check — was blocking channels when `resolveSlackChannelConfig()` returned `{ allowed: false }` for unlisted channels
The second check conflated "channel not in config" with "channel explicitly denied."
## Fix
Use `matchSource` to distinguish explicit denial from absence of config:
```ts
const hasExplicitConfig = Boolean(channelConfig?.matchSource);
if (!channelAllowed && (params.groupPolicy !== "open" || hasExplicitConfig)) {
return false;
}
```
When `matchSource` is undefined, the channel has no explicit config entry and should be allowed under open policy.
## Behavior After Fix
| Scenario | Result |
|----------|--------|
| `groupPolicy: "open"`, channel unlisted | ✅ Allowed |
| `groupPolicy: "open"`, channel explicitly denied (`allow: false`) | ❌ Blocked |
| `groupPolicy: "open"`, channel with custom config | ✅ Allowed |
| `groupPolicy: "allowlist"`, channel unlisted | ❌ Blocked |
## Test Plan
- [x] Open policy + unlisted channel → allowed
- [x] Open policy + explicitly denied channel → blocked
- [x] Allowlist policy + unlisted channel → blocked
- [x] Allowlist policy + listed channel → allowed
Added a dedicated Anthropic payload logger that writes exact request
JSON (as sent) plus per‑run usage stats (input/output/cache read/write)
to a
standalone JSONL file, gated by an env flag.
Changes
- New logger: src/agents/anthropic-payload-log.ts (writes
logs/anthropic-payload.jsonl under the state dir, optional override via
env).
- Hooked into embedded runs to wrap the stream function and record
usage: src/agents/pi-embedded-runner/run/attempt.ts.
How to enable
- CLAWDBOT_ANTHROPIC_PAYLOAD_LOG=1
- Optional:
CLAWDBOT_ANTHROPIC_PAYLOAD_LOG_FILE=/path/to/anthropic-payload.jsonl
What you’ll get (JSONL)
- stage: "request" with payload (exact Anthropic params) +
payloadDigest
- stage: "usage" with usage
(input/output/cacheRead/cacheWrite/totalTokens/etc.)
Notes
- Usage is taken from the last assistant message in the run; if the
run fails before usage is present, you’ll only see an error field.
Files touched
- src/agents/anthropic-payload-log.ts
- src/agents/pi-embedded-runner/run/attempt.ts
Tests not run.
- Add registry lock during command execution to prevent race conditions
- Add input sanitization for command arguments (defense in depth)
- Validate handler is a function during registration
- Remove redundant case-insensitive regex flag
- Add success logging for command execution
- Simplify handler return type (always returns result now)
- Remove dead code branch in commands-plugin.ts
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Add clearPluginCommands() call in loadClawdbotPlugins() to ensure
previously registered commands are cleaned up before reloading plugins.
This prevents command conflicts during hot-reload scenarios.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Blockers fixed:
- Fix documentation: requireAuth defaults to true (not false)
- Add command name validation (must start with letter, alphanumeric only)
- Add reserved commands list to prevent shadowing built-in commands
- Emit diagnostic errors for invalid/duplicate command registration
Other improvements:
- Return user-friendly message for unauthorized commands (instead of silence)
- Sanitize error messages to avoid leaking internal details
- Document acceptsArgs behavior when arguments are provided
- Add notes about reserved commands and validation rules to docs
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
This adds a new `api.registerCommand()` method to the plugin API, allowing
plugins to register slash commands that execute without invoking the AI agent.
Features:
- Plugin commands are processed before built-in commands and the agent
- Commands can optionally require authorization
- Commands can accept arguments
- Async handlers are supported
Use case: plugins can implement toggle commands (like /tts_on, /tts_off)
that respond immediately without consuming LLM API calls.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
* feat: skip heartbeat API calls when HEARTBEAT.md is effectively empty
- Added isHeartbeatContentEffectivelyEmpty() to detect files with only headers/comments
- Modified runHeartbeatOnce() to check HEARTBEAT.md content before polling the LLM
- Returns early with 'empty-heartbeat-file' reason when no actionable tasks exist
- Preserves existing behavior when file is missing (lets LLM decide)
- Added comprehensive test coverage for empty file detection
- Saves API calls/costs when heartbeat file has no meaningful content
* chore: update HEARTBEAT.md template to be effectively empty by default
Changed instruction text to comment format so new workspaces benefit from
heartbeat optimization immediately. Users still get clear guidance on usage.
* fix: only treat markdown headers (# followed by space) as comments, not #TODO etc
* refactor: simplify regex per code review suggestion
* docs: clarify heartbeat empty file behavior (#1535) (thanks @JustYannicc)
---------
Co-authored-by: Peter Steinberger <steipete@gmail.com>
Add automatic discovery of AWS Bedrock models using ListFoundationModels API.
When AWS credentials are detected, models that support streaming and text output
are automatically discovered and made available.
- Add @aws-sdk/client-bedrock dependency
- Add discoverBedrockModels() with caching (default 1 hour)
- Add resolveImplicitBedrockProvider() for auto-registration
- Add BedrockDiscoveryConfig for optional filtering by provider/region
- Filter to active, streaming, text-output models only
- Update docs/bedrock.md with auto-discovery documentation
The left navigation sidebar now stays fixed when scrolling through
long content pages like /skills. Changed .shell from min-height to
fixed height with overflow: hidden, allowing nav and content to
scroll independently within their grid cells.
Co-authored-by: pookNast <pook@nast.local>
Co-authored-by: Claude Opus 4.5 <noreply@anthropic.com>
- New docs/automation/cron-vs-heartbeat.md with complete guidance
- Cross-links from heartbeat.md and cron-jobs.md
- Updated AGENTS.md template with practical guidance
- Added navigation entry in docs.json
Updated `dockerImageExists` in `src/commands/doctor-sandbox.ts` to mirror the logic in `src/agents/sandbox/docker.ts`. It now re-throws errors unless they are explicitly "No such image" errors.
Previously, `dockerImageExists` assumed any error from `docker image inspect` meant the image did not exist. This masked other errors like socket permission issues.
This change:
- Modifies `dockerImageExists` to inspect stderr when the exit code is non-zero.
- Returns `false` only if the error explicitly indicates "No such image" or "No such object".
- Throws an error with the stderr content for all other failures.
- Adds a reproduction test in `src/agents/sandbox/docker.test.ts`.
Adds built-in Tlon (Urbit) channel plugin to support decentralized messaging on the Urbit network.
Features:
- DM and group chat support
- SSE-based real-time message monitoring
- Auto-discovery of group channels
- Thread replies and reactions
- Integration with Urbit's HTTP API
This resolves cron delivery issues with external Tlon plugins by making it a first-class built-in channel alongside Telegram, Signal, and other messaging platforms.
Implementation includes:
- Plugin registration via ClawdbotPluginApi
- Outbound delivery with sendText and sendMedia
- Gateway adapter for inbound message handling
- Urbit SSE client for event streaming
- Core bridge for Clawdbot runtime integration
Co-authored-by: William Arzt <william@arzt.co>
To avoid permission denied errors when modifying Tailscale configuration (serve/funnel),
we now attempt the command directly first. If it fails, we catch the error and retry
with `sudo -n`. This preserves existing behavior for users where it works, but
attempts to escalate privileges (non-interactively) if needed.
- Added `execWithSudoFallback` helper in `src/infra/tailscale.ts`.
- Updated `ensureFunnel`, `enableTailscaleServe`, `disableTailscaleServe`,
`enableTailscaleFunnel`, and `disableTailscaleFunnel` to use the fallback helper.
- Added tests in `src/infra/tailscale.test.ts` to verify fallback behavior.
To avoid permission denied errors when modifying Tailscale configuration (serve/funnel),
we now prepend `sudo -n` to these commands. This ensures that if the user has appropriate
sudo privileges (specifically passwordless for these commands or generally), the operation
succeeds. If sudo fails (e.g. requires password non-interactively), it will throw an error
which is caught and logged as a warning, preserving existing behavior but attempting to escalate privileges first.
- Updated `ensureFunnel` to use `sudo -n` for the enabling step.
- Updated `enableTailscaleServe`, `disableTailscaleServe`, `enableTailscaleFunnel`, `disableTailscaleFunnel` to use `sudo -n`.
To avoid permission denied errors when modifying Tailscale configuration (serve/funnel),
we now prepend `sudo -n` to these commands. This ensures that if the user has appropriate
sudo privileges (specifically passwordless for these commands or generally), the operation
succeeds. If sudo fails (e.g. requires password non-interactively), it will throw an error
which is caught and logged as a warning, preserving existing behavior but attempting to escalate privileges first.
- Updated `ensureFunnel` to use `sudo -n` for the enabling step.
- Updated `enableTailscaleServe`, `disableTailscaleServe`, `enableTailscaleFunnel`, `disableTailscaleFunnel` to use `sudo -n`.
- Added tests in `src/infra/tailscale.test.ts` to verify `sudo` usage.
- Set inputTokens, outputTokens, totalTokens to 0 in sessions.reset
- Clear TUI sessionInfo tokens immediately before async reset
- Prevents stale token display after session reset
Fixes#1523
The @microsoft/agents-hosting SDK's MsalTokenProvider automatically
appends `/.default` to all scope strings in its token acquisition
methods (acquireAccessTokenViaSecret, acquireAccessTokenViaFIC,
acquireAccessTokenViaWID, acquireTokenWithCertificate in
msalTokenProvider.ts). This is consistent SDK behavior, not a recent
change.
Our code was including `.default` in scope URLs, resulting in invalid
double suffixes like `https://graph.microsoft.com/.default/.default`.
This was confirmed to cause Graph API authentication errors. Removing
the `.default` suffix from our scope strings allows the SDK to append
it correctly, resolving the issue.
Before: we pass `.default` -> SDK appends -> double `.default` (broken)
After: we pass base URL -> SDK appends -> single `.default` (works)
Co-authored-by: Christof Salis <c.salis@vertifymed.com>
Adds built-in Tlon (Urbit) channel plugin to support decentralized messaging on the Urbit network.
Features:
- DM and group chat support
- SSE-based real-time message monitoring
- Auto-discovery of group channels
- Thread replies and reactions
- Integration with Urbit's HTTP API
This resolves cron delivery issues with external Tlon plugins by making it a first-class built-in channel alongside Telegram, Signal, and other messaging platforms.
Implementation includes:
- Plugin registration via ClawdbotPluginApi
- Outbound delivery with sendText and sendMedia
- Gateway adapter for inbound message handling
- Urbit SSE client for event streaming
- Core bridge for Clawdbot runtime integration
Co-authored-by: William Arzt <william@arzt.co>
Fixed issue where trigger words would disappear when typing or when adding new trigger words. The problem was that `swabbleTriggerWords` changes were triggering `VoiceWakeRuntime.refresh()` which sanitized the array by removing empty strings in real-time.
Solution: Introduced local `@State` buffer `triggerEntries` with stable UUID identifiers for each trigger word entry. User edits now only affect the local state buffer and are synced back to `AppState` on explicit actions (submit, remove, disappear). This prevents premature sanitization during editing.
The local state is loaded on view appear and when the view becomes active, ensuring it stays in sync with `AppState`.
* fix(discord): gate autoThread by thread owner
* fix(discord): ack bot-owned autoThreads
* fix(discord): ack mentions in open channels
- Ack reactions in bot-owned autoThreads
- Ack reactions in open channels (no mention required)
- DRY: Pass pre-computed isAutoThreadOwnedByBot to avoid redundant checks
- Consolidate ack logic with explanatory comment
* fix: allow null values in exec.approval.request schema
The ExecApprovalRequestParamsSchema was rejecting null values for optional
fields like resolvedPath, but the calling code in bash-tools.exec.ts passes
null. This caused intermittent 'invalid exec.approval.request params'
validation errors.
Fix: Accept Type.Union([Type.String(), Type.Null()]) for all optional string
fields in the schema. Update test to reflect new behavior.
* fix: align discord ack reactions with mention gating (#1511) (thanks @pvoo)
---------
Co-authored-by: Wimmie <wimmie@tameson.com>
Co-authored-by: Peter Steinberger <steipete@gmail.com>
* fix(linux): add user bin directories to systemd service PATH
Fixes#1503
On Linux, the systemd service PATH was hardcoded to only include system
directories (/usr/local/bin, /usr/bin, /bin), causing binaries installed
via npm global with custom prefix or node version managers to not be found.
This adds common Linux user bin directories to the PATH:
- ~/.local/bin (XDG standard, pip, etc.)
- ~/.npm-global/bin (npm custom prefix)
- ~/bin (user's personal bin)
- Node version manager paths (nvm, fnm, volta, asdf)
- ~/.local/share/pnpm (pnpm global)
- ~/.bun/bin (Bun)
User directories are added before system directories so user-installed
binaries take precedence.
🤖 AI-assisted (Claude Opus 4.5 via Clawdbot)
📋 Testing: Existing unit tests pass (7/7)
* test: add comprehensive tests for Linux user bin directory resolution
- Add dedicated tests for resolveLinuxUserBinDirs() function
- Test path ordering (extraDirs > user dirs > system dirs)
- Test buildMinimalServicePath() with HOME set/unset
- Test platform-specific behavior (Linux vs macOS vs Windows)
Test count: 7 → 20 (+13 tests)
* test: add comprehensive tests for Linux user bin directory handling
- Test Linux user directories included when HOME is set
- Test Linux user directories excluded when HOME is missing
- Test path ordering (extraDirs > user dirs > system dirs)
- Test platform-specific behavior (Linux vs macOS vs Windows)
- Test buildMinimalServicePath() with HOME in env
Covers getMinimalServicePathParts() and buildMinimalServicePath()
for all Linux user bin directory edge cases.
Test count: 7 → 16 (+9 tests)
Tables render poorly in Telegram (pipes stripped, whitespace collapses).
This adds a 'tableMode' option to markdownToIR that converts tables to
nested bullet points, which render cleanly on mobile.
- Add tableMode: 'flat' | 'bullets' to MarkdownParseOptions
- Track table state during token rendering
- Render tables as bullet points with first column as row labels
- Apply bold styling to row labels for visual hierarchy
- Enable tableMode: 'bullets' for Telegram formatter
Closes #TBD
When installing the LaunchAgent/systemd service, the CLI was using
fs.realpath() to resolve the entry.js path, which converted stable
symlinked paths (e.g. node_modules/clawdbot) into version-specific
paths (e.g. .pnpm/clawdbot@X.Y.Z/...).
This caused the service to break after pnpm updates because the old
versioned path no longer exists, even though the symlink still works.
Now we prefer the original (symlinked) path when it's valid, keeping
service configs stable across package version updates.
Adds sanitization to extractAssistantText in sessions-helpers.ts to
prevent tool call text from leaking to users. Previously, messages
retrieved from chat history via sessions-helpers.ts could expose:
- Minimax XML tool calls (<invoke>...</invoke>)
- Downgraded tool call markers ([Tool Call: name (ID: ...)])
- Thinking tags (<think>...</think>)
This fix:
- Exports the stripping functions from pi-embedded-utils.ts
- Adds a new sanitizeTextContent helper in sessions-helpers.ts
- Updates extractAssistantText to sanitize before returning
- Updates extractMessageText in commands-subagents.ts to sanitize
Fixes#1269
Co-Authored-By: Claude <noreply@anthropic.com>
* fix(ui): allow relative URLs in avatar validation
The isAvatarUrl check only accepted http://, https://, or data: URLs,
but the /avatar/{agentId} endpoint returns relative paths like /avatar/main.
This caused local file avatars to display as text instead of images.
Fixes avatar display for locally configured avatar files.
* fix(gateway): resolve local avatars to URL in HTML injection and RPC
The frontend fix alone wasn't enough because:
1. serveIndexHtml() was injecting the raw avatar filename into HTML
2. agent.identity.get RPC was returning raw filename, overwriting the
HTML-injected value
Now both paths resolve local file avatars (*.png, *.jpg, etc.) to the
/avatar/{agentId} endpoint URL.
* feat(compaction): add adaptive chunk sizing and progressive fallback
- Add computeAdaptiveChunkRatio() to reduce chunk size for large messages
- Add isOversizedForSummary() to detect messages too large to summarize
- Add summarizeWithFallback() with progressive fallback:
- Tries full summarization first
- Falls back to partial summarization excluding oversized messages
- Notes oversized messages in the summary output
- Add SAFETY_MARGIN (1.2x) buffer for token estimation inaccuracy
- Reduce MIN_CHUNK_RATIO to 0.15 for very large messages
This prevents compaction failures when conversations contain
unusually large tool outputs or responses that exceed the
summarization model's context window.
* feat(ui): add compaction indicator and improve event error handling
Compaction indicator:
- Add CompactionStatus type and handleCompactionEvent() in app-tool-stream.ts
- Show '🧹 Compacting context...' toast while active (with pulse animation)
- Show '🧹 Context compacted' briefly after completion
- Auto-clear toast after 5 seconds
- Add CSS styles for .callout.info, .callout.success, .compaction-indicator
Error handling improvements:
- Wrap onEvent callback in try/catch in gateway.ts to prevent errors
from breaking the WebSocket message handler
- Wrap handleGatewayEvent in try/catch with console.error logging
to isolate errors and make them visible in devtools
These changes address UI freezes during heavy agent activity by:
1. Showing users when compaction is happening
2. Preventing uncaught errors from silently breaking the event loop
* fix(control-ui): add agentId to DEFAULT_ASSISTANT_IDENTITY
TypeScript inferred the union type without agentId when falling back to
DEFAULT_ASSISTANT_IDENTITY, causing build errors at control-ui.ts:222-223.
Adds support for separate replyToMode settings for DMs vs channels:
- Add channels.slack.dm.replyToMode for DM-specific threading
- Keep channels.slack.replyToMode as default for channels
- Add resolveSlackReplyToMode helper to centralize logic
- Pass chatType through threading resolution chain
Usage:
```json5
{
channels: {
slack: {
replyToMode: "off", // channels
dm: {
replyToMode: "all" // DMs always thread
}
}
}
}
```
When dm.replyToMode is set, DMs use that mode; channels use the
top-level replyToMode. Backward compatible when not configured.
When replying to a Slack thread, files attached to the root message were
not being fetched. The existing `resolveSlackThreadStarter()` fetched the
root message text via `conversations.replies` but ignored the `files[]`
array in the response.
Changes:
- Add `files` to `SlackThreadStarter` type and extract from API response
- Download thread starter files when the reply message has no attachments
- Add verbose log for thread starter file hydration
Fixes issue where asking about a PDF in a thread reply would fail because
the model never received the file content from the root message.
Add asVoice parameter to sendBlueBubblesAttachment that converts audio
to iMessage voice memo format (Opus CAF at 48kHz) and sets isAudioMessage
flag in the BlueBubbles API.
This follows the existing asVoice pattern used by Telegram.
- Convert audio to Opus CAF format using ffmpeg when asVoice=true
- Set isAudioMessage=true in BlueBubbles attachment API
- Pass asVoice through action handler and media-send
- Add regex caching to avoid creating new RegExp objects on each render
- Optimize smartFilter to use single array with tier-based scoring
- Replace non-existent fuzzyFilter import with local fuzzyFilterLower
- Reduces from 4 array allocations and 4 sorts to 1 array and 1 sort
Fixes pre-existing bug where fuzzyFilter was imported from pi-tui but not exported.
Auth profiles in cooldown (due to rate limiting) were being attempted,
causing unnecessary retries and delays. This fix ensures:
1. Initial profile selection skips profiles in cooldown
2. Profile rotation (after failures) skips cooldown profiles
3. Clear error message when all profiles are unavailable
Tests added:
- Skips profiles in cooldown during initial selection
- Skips profiles in cooldown when rotating after failure
Fixes#1316
The progress spinner was being shown for each gateway RPC call during
log tailing, causing repeated spinner frames (◇ │) to appear every
polling interval.
Add a `progress` option to `callGatewayFromCli` and disable the spinner
during follow mode polling to keep output clean.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Previously, typing stop was intentionally skipped because the
BlueBubbles Server DELETE endpoint was bugged (called startTyping
instead of stopTyping). Now that the server bug is fixed, we can
properly stop typing indicators.
- onIdle: now calls sendBlueBubblesTyping(false) to stop typing
- finally block: stops typing when no message sent (NO_REPLY case)
Tests verify:
- Success message shown when session state available
- Error message shown when sessionEntry missing
- Error message shown when sessionStore missing
- No model message when no /model directive
Covers edge cases for #1435 fix.
Previously, the /model command would display 'Model set to X' even when
the session state wasn't actually persisted (when sessionEntry, sessionStore,
or sessionKey were missing). This caused confusion as users saw success
messages but the model didn't actually change.
This fix:
- Tracks whether the model override was actually persisted
- Only shows success message when persist happened
- Shows a clear error message when persist fails
AI-assisted: Claude Opus 4.5 via Clawdbot
Testing: lightly tested (code review, no runtime test)
The session-memory hook saves session context to memory files when /new is run,
which is useful internal housekeeping. However, the confirmation message that
was displayed to users (showing the file path) leaked implementation details.
This change removes the user-visible message while keeping the console.log
for debugging purposes. The hook continues to save session context silently.
The frontend fix alone wasn't enough because:
1. serveIndexHtml() was injecting the raw avatar filename into HTML
2. agent.identity.get RPC was returning raw filename, overwriting the
HTML-injected value
Now both paths resolve local file avatars (*.png, *.jpg, etc.) to the
/avatar/{agentId} endpoint URL.
The isAvatarUrl check only accepted http://, https://, or data: URLs,
but the /avatar/{agentId} endpoint returns relative paths like /avatar/main.
This caused local file avatars to display as text instead of images.
Fixes avatar display for locally configured avatar files.
Move mattermost channel implementation from core to extensions/mattermost plugin. Extract config schema, group mentions, normalize utilities, and all mattermost-specific logic (accounts, client, monitor, probe, send) into the extension. Update imports to use plugin SDK and local modules. Add channel metadata directly in plugin definition instead of using getChatChannelMeta. Update package.json with channel and install configuration.
Replace invalid "docs" service with the correct "tasks,people" services
in the setup example. The gog CLI does not have a "docs" service - docs
commands (export/cat) use Drive authentication instead.
Fixes#1433
Co-Authored-By: Claude <noreply@anthropic.com>
Slack's files.uploadV2 API no longer supports the filetype field and logs
deprecation warnings when it's included. Slack auto-detects the file type
from the file content, so this field is unnecessary.
This removes the warning:
[WARN] web-api:WebClient filetype is no longer a supported field in files.uploadV2.
The message tool accepts path and filePath parameters in its schema,
but these were never converted to mediaUrl, causing local files to
be ignored when sending messages.
Changes:
- src/agents/tools/message-tool.ts: Convert path/filePath to media with file:// URL
- src/infra/outbound/message-action-runner.ts: Allow hydrateSendAttachmentParams for "send" action
Fixes issue where local audio files (and other media) couldn't be sent
via the message tool with the path parameter.
Users can now use:
message({ path: "/tmp/file.ogg" })
message({ filePath: "/tmp/file.ogg" })
- Add regex caching to avoid creating new RegExp objects on each render
- Optimize smartFilter to use single array with tier-based scoring
- Replace non-existent fuzzyFilter import with local fuzzyFilterLower
- Reduces from 4 array allocations and 4 sorts to 1 array and 1 sort
Fixes pre-existing bug where fuzzyFilter was imported from pi-tui but not exported.
Auth profiles in cooldown (due to rate limiting) were being attempted,
causing unnecessary retries and delays. This fix ensures:
1. Initial profile selection skips profiles in cooldown
2. Profile rotation (after failures) skips cooldown profiles
3. Clear error message when all profiles are unavailable
Tests added:
- Skips profiles in cooldown during initial selection
- Skips profiles in cooldown when rotating after failure
Fixes#1316
Adds the ability to customize the assistant's name and avatar in the Web UI.
Configuration options:
- config.ui.assistant.name: Custom name (replaces 'Assistant')
- config.ui.assistant.avatar: Emoji or letter for avatar (replaces 'A')
Also reads from workspace IDENTITY.md as fallback:
- Name: field sets the assistant name
- Emoji: field sets the avatar
Priority: config > IDENTITY.md > defaults
Closes#1383
Export the SECTION_META constant from config-form.render.ts and
re-export it through config-form.ts so it can be imported by config.ts.
This fixes a runtime error where SECTION_META was being referenced
but not properly exported from its source module.
When creating exec tools via chat/Discord, agentId was not passed,
causing allowlist lookup to use 'default' key instead of 'main'.
User's allowlist entries in agents.main were never matched.
Now derives agentId from sessionKey if not explicitly provided,
ensuring correct allowlist lookup for all exec paths.
- Add avatar field to IdentityConfig type
- Add avatar parsing in AgentIdentity from IDENTITY.md
- Add renderAvatar support for image avatars in webchat
- Add CSS styling for image avatars
Users can now configure a custom avatar for the assistant in the webchat
by setting 'identity.avatar' in the agent config or adding 'Avatar: path'
to IDENTITY.md. The avatar can be served from the assets folder.
Closes #TBD
TypeBox Type.Optional(Type.String()) accepts string|undefined but NOT null.
Discord exec was failing with 'resolvedPath must be string' because callers
passed null explicitly. Web UI worked because it skipped the approval request.
Fixes exec approval validation error in Discord-triggered sessions.
The debug view now automatically refreshes every 3 seconds when active,
similar to the logs view. This removes the need to manually click the
refresh button to see updated debug messages and status information.
- Add ToolCallIdMode type ('standard' | 'strict') for provider compatibility
- Standard mode (default): allows [a-zA-Z0-9_-] for readable session logs
- Strict mode: only [a-zA-Z0-9] for Mistral via OpenRouter
- Update sanitizeSessionMessagesImages to accept toolCallIdMode option
- Export ToolCallIdMode from pi-embedded-helpers barrel
Addresses review feedback on PR #1372 about readability.
Some providers like Mistral via OpenRouter require strictly alphanumeric
tool call IDs. The error message indicates: "Tool call id was
whatsapp_login_1768799841527_1 but must be a-z, A-Z, 0-9, with a length
of 9."
Changes:
- Update sanitizeToolCallId to strip all non-alphanumeric characters
(previously allowed underscores and hyphens)
- Update makeUniqueToolId to use alphanumeric suffixes (x2, x3, etc.)
instead of underscores
- Update isValidCloudCodeAssistToolId to validate alphanumeric-only IDs
- Update tests to reflect stricter sanitization
Fixes#1359
Co-Authored-By: Claude <noreply@anthropic.com>
Closes#1393
The distribution script (package-mac-dist.sh) now defaults BUILD_ARCHS to 'all',
producing universal binaries that run natively on both Apple Silicon and Intel Macs.
Previously, the script inherited the host architecture default from package-mac-app.sh,
which meant release builds done on ARM Macs only included ARM binaries.
Closes#1393
The distribution script (package-mac-dist.sh) now defaults BUILD_ARCHS to 'all',
producing universal binaries that run natively on both Apple Silicon and Intel Macs.
Previously, the script inherited the host architecture default from package-mac-app.sh,
which meant release builds done on ARM Macs only included ARM binaries.
Add Mattermost as a supported messaging channel with bot API and WebSocket integration. Includes channel state tracking (tint, summary, details), multi-account support, and delivery target routing. Update documentation and tests to include Mattermost alongside existing channels.
* feat(sessions): add channelIdleMinutes config for per-channel session idle durations
Add new `channelIdleMinutes` config option to allow different session idle
timeouts per channel. For example, Discord sessions can now be configured
to last 7 days (10080 minutes) while other channels use shorter defaults.
Config example:
sessions:
channelIdleMinutes:
discord: 10080 # 7 days
The channel-specific idle is passed as idleMinutesOverride to the existing
resolveSessionResetPolicy, integrating cleanly with the new reset policy
architecture.
* fix
* feat: add per-channel session reset overrides (#1353) (thanks @cash-echo-bot)
---------
Co-authored-by: Cash Williams <cashwilliams@gmail.com>
Co-authored-by: Peter Steinberger <steipete@gmail.com>
Add support for '*' wildcard in Discord channel configuration,
matching the existing guild-level wildcard behavior.
This allows applying default channel settings (like autoThread)
to all channels without listing each one explicitly:
guilds:
'*':
channels:
'*': { autoThread: true }
Specific channel configs still take precedence over the wildcard.
Added a standalone cache tracing module and wired it into the embedded
runner so you can capture message flow and the exact context sent to
Anthropic in a separate JSONL file.
What changed
- New tracing module: src/agents/cache-trace.ts (self‑contained,
env‑gated, writes JSONL, computes per‑message digests).
- Hook points in src/agents/pi-embedded-runner/run/attempt.ts: logs
stage snapshots (loaded/sanitized/limited/prompt/stream/after) and wraps
the
stream fn to record the real context.messages at send time.
How to enable
- CLAWDBOT_CACHE_TRACE=1 enables tracing.
- CLAWDBOT_CACHE_TRACE_FILE=~/.clawdbot/logs/cache-trace.jsonl
overrides output (default is
$CLAWDBOT_STATE_DIR/logs/cache-trace.jsonl).
- Optional filters:
- CLAWDBOT_CACHE_TRACE_MESSAGES=0 to omit full messages (still
logs digests).
- CLAWDBOT_CACHE_TRACE_PROMPT=0 to omit prompt text.
- CLAWDBOT_CACHE_TRACE_SYSTEM=0 to omit system prompt.
What you’ll see
- One JSON object per line with stage, messagesDigest, per‑message
messageFingerprints, and the actual messages if enabled.
- The most important line is stage: "stream:context" — that is the
exact payload pi‑mono is sending. If this diverges from earlier stages,
you’ve
found the mutation point.
- Updated message processing to include full message IDs alongside short IDs for better context resolution.
- Improved reply handling by caching inbound messages, allowing for accurate sender and body resolution without exposing dropped content.
- Adjusted tests to validate the new full ID properties and their integration into the message handling workflow.
- Implemented resolution of short message IDs to full UUIDs in both text and media sending functions.
- Updated reply context formatting to optimize token usage by including only necessary information.
- Introduced truncation for long reply bodies to further reduce token consumption.
- Adjusted tests to reflect changes in reply context handling and message ID resolution.
- Added functionality to resolve short message IDs to full UUIDs and vice versa, optimizing token usage.
- Introduced a reply cache to store message context for replies when metadata is omitted in webhook payloads.
- Updated message handling to utilize short IDs for outbound messages and replies, improving efficiency.
- Enhanced error messages to clarify required parameters for actions like react, edit, and unsend.
- Added tests to ensure correct behavior of new features and maintain existing functionality.
- Add per-message dedup tracking in subscribeEmbeddedPiSession
- Compare both trimmed and normalized text to catch near-duplicates
- Reset dedup state on each new assistant message
- Add test for trailing whitespace edge case
Fixes duplicate Slack message delivery when the same text appears
with minor whitespace differences (e.g., trailing newline).
Fixes two bugs in Slack tool notification delivery:
1. Tool notifications ignored verbose=false - normalized verbose values so
boolean false/'false' are properly treated as 'off'
2. Thread context lost - Slack outbound adapter now falls back to threadId
when replyToId is missing, and MessageThreadId is set for thread replies
Closes#1333
The iOS app currently identifies as clientId=clawdbot-ios when
connecting in node mode. Add this ID to the allowed gateway client
IDs so the handshake schema accepts it.
Also fixes a TS strictness issue in auto-reply status formatting
(parts filter) that caused
> clawdbot@2026.1.20 build /Users/vignesh/clawd/clawdbot-upstream
> tsc -p tsconfig.json && node --import tsx scripts/canvas-a2ui-copy.ts && node --import tsx scripts/copy-hook-metadata.ts && node --import tsx scripts/write-build-info.ts
[copy-hook-metadata] Copied boot-md/HOOK.md
[copy-hook-metadata] Copied command-logger/HOOK.md
[copy-hook-metadata] Copied session-memory/HOOK.md
[copy-hook-metadata] Copied soul-evil/HOOK.md
[copy-hook-metadata] Done to fail.
- Disable Talk mode start on iOS simulator (no audio input)
- Validate audio input format before installing tap to avoid
AVFAudio assertion crashes on misconfigured devices.
Tested:
- Launched app on iOS simulator and tapping Talk no longer crashes
(shows error path instead).
Use the timeout provided on node invoke requests to ensure node
clients always respond with a result.
This prevents gateway-side node.invoke calls from hanging until the
gateway timeout when a node command stalls.
Tests:
- swift test --filter GatewayNodeSessionTests
Record submitted inputs in the editor history so up/down arrow
can recall previous messages.
Adds a small helper to wire submit handling and unit tests for
routing/recording behavior.
No PR yet (per request).
Telegram retries callback queries if they aren't acknowledged quickly.
Previously, answerCallbackQuery was called in a finally block AFTER
processing, which could take several seconds for agent responses.
This change moves answerCallbackQuery to immediately after basic
validation, before any processing begins. This prevents Telegram
from sending duplicate callbacks while the agent is thinking.
Fixes duplicate callback handling when agent processing is slow.
Improve validation error formatting for strict schemas.
In particular, additionalProperties errors now surface the
unexpected property name and where it occurred, which makes
handshake/connect failures easier to debug.
- Add failing tests for delivery field preservation
- Fix register.cron-edit to conditionally build payload object
- Only include delivery fields (deliver, channel, to, bestEffortDeliver) when explicitly provided
- Previously undefined values were included, wiping out existing delivery settings
- Now --message alone preserves existing delivery config
- Tests verify both preservation and explicit override scenarios
Adds support for passing `gatewayUrl` as a URL parameter to the WebChat UI,
allowing the control-ui to connect to a remote gateway (e.g., VPS) instead
of defaulting to localhost.
Usage: http://localhost:5173/?gatewayUrl=ws://<vps-ip>:18789&token=<token>
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Fixes#1327 - 'clawdbot message --channel telegram' fails with
'Unknown channel: telegram' because plugins weren't loaded.
The Commander code path (non-route-first) calls ensureConfigReady() in
preAction but doesn't load the plugin registry. Channel plugins like
telegram are registered during plugin loading, so getChannelPlugin()
returns undefined without it.
This adds ensurePluginRegistryLoaded() call for commands that need
channel plugin access: message, channels, directory.
- Add avatar field to IdentityConfig type
- Add avatar parsing in AgentIdentity from IDENTITY.md
- Add renderAvatar support for image avatars in webchat
- Add CSS styling for image avatars
Users can now configure a custom avatar for the assistant in the webchat
by setting 'identity.avatar' in the agent config or adding 'Avatar: path'
to IDENTITY.md. The avatar can be served from the assets folder.
Closes #TBD
Move dynamic import of @mariozechner/pi-coding-agent into the try/catch so transient module resolution errors don't poison the model catalog cache with a rejected promise.
This previously caused Discord/Telegram handlers and heartbeat to fail until process restart if the import failed once.
The device identity check was rejecting connections before token
authentication could be attempted. This broke the control-ui (web UI)
which uses token-based authentication via URL parameter.
Changes:
- Skip device identity requirement when a token is provided
- Guard device token verification to only run when device is present
Fixes control-ui showing "device identity required" error when
connecting with a valid token.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Sidebar:
- SVG icons instead of emoji (consistent rendering)
- Clean navigation with active states
Form fields completely redesigned:
- Toggle rows: full-width clickable with label + description
- Segmented controls: for enum values with ≤5 options
- Number inputs: with +/- stepper buttons
- Text inputs: with reset-to-default button
- Select dropdowns: clean styling with custom arrow
- Arrays: card-based with clear add/remove, item numbering
- Objects: collapsible sections with chevron animation
- Maps: key-value editor with inline editing
Visual improvements:
- Consistent border radius and spacing
- Better color contrast for labels vs help text
- Hover and focus states throughout
- Icons for common actions (add, remove, reset)
Mobile:
- Horizontal scrolling nav on small screens
- Stacked layouts for complex fields
- Extract formatRelativeTime to shared utility for reuse across components
- Optimize FilterableSelectList with pre-lowercased searchTextLower field (avoids toLowerCase on every keystroke)
- Implement custom fuzzy matching with space-separated token support and word boundary scoring
- Use matchesKey utility for consistent keybinding handling (arrows, vim j/k, ctrl+p/n)
- Fix searchable-select-list to support vim keybindings consistently
- Fix system-prompt runtimeInfo null check with nullish coalescing operator
Read the final user/assistant message from session transcripts and display
it in the picker alongside the session update time. Allows quick previews
of what's in each session without opening it.
Replace hardcoded escape sequence checks with the pi-tui keybindings API to ensure consistent cancel handling across different terminal configurations.
Server-side filtering backup for client-side session picker search.
Case-insensitive substring match on displayName, label, subject,
sessionId, and key.
Closes#1161
- Remove dead input handlers (onSubmit/onEscape never triggered)
- Store maxVisible as instance property instead of bracket access
- Remove unused filterInput theme property
Users can now type to filter sessions in real-time:
- FilterableSelectList component wraps pi-tui's fuzzyFilter
- Matches against displayName, label, subject, sessionId
- j/k navigation, Enter selects, Escape clears filter then cancels
- Uses derivedTitle from previous commit for better display
Refs #1161
Enable meaningful session titles via priority-based derivation:
1. displayName (user-set)
2. subject (group name)
3. First user message (truncated to 60 chars)
4. sessionId prefix + date fallback
Opt-in via includeDerivedTitles param to avoid perf impact on
regular listing. Reads only first 10 lines of transcript files.
Closes#1161
Major redesign of the config page:
Layout:
- Sidebar navigation with section list
- Search input to filter settings
- Section cards with icons and descriptions
- Responsive design for mobile (stacked layout)
Fields:
- New toggle switches for booleans (replaces checkboxes)
- Improved field-row layout with label, help text, and control
- Better fieldset and array styling
Features:
- Diff view showing pending changes before save
- Original value tracking for comparison
- Section filtering via sidebar nav
- Search across setting names, descriptions, and nested properties
Styling:
- Dedicated config.css with all new styles
- Dark and light theme support
- Smooth animations and transitions
- Mobile-first responsive breakpoints
- Group config settings into logical sections (Core, Agents, Communication, etc.)
- Add collapsible accordion UI for each section group
- Add icons and labels for each config category
- Improve mobile responsiveness with better button layout
- Style improvements for nested fieldsets and arrays
When using the default AI snapshot format without explicit options like
interactive/compact/labels, refs were not being registered because
snapshotAiViaPlaywright returns raw text without ref registration.
This caused 'Unknown ref' errors when subsequently using act commands
with refs like e12 that appeared in the snapshot text.
The fix extracts refs from the AI snapshot using buildRoleSnapshotFromAiSnapshot
and registers them via rememberRoleRefsForTarget so act commands can resolve them.
Fixes#1268
The regex `/context window/i` was matching "Model context window too small"
and rewriting it to generic "Context overflow" message, hiding the actual
problem from users.
Add exclusion for "too small|minimum is" patterns so the original
informative error message passes through.
🤖 AI-assisted (Claude) - tested on local instance
Co-Authored-By: Claude <noreply@anthropic.com>
Fixes#1172
- Add per-runId message buffering in ChatLog
- Separate thinking stream from content stream handling
- Ensure proper sequencing (thinking always before content)
- Model-agnostic: works with or without thinking tokens
- Render assistant reasoning as a distinct block (not merged into message text).\n- Detect tool-like messages reliably and style them separately.\n- Add a "🧠" toggle to hide/show tool + thinking output, persisted in UI settings.
- Add input_image and input_file support with SSRF protection
- Add client-side tools (Hosted Tools) support
- Add turn-based tool flow with function_call_output handling
- Export buildAgentPrompt for testing
Add a new `/v1/responses` endpoint implementing the OpenResponses API
standard for agentic workflows. This provides:
- Item-based input (messages, function_call_output, reasoning)
- Semantic streaming events (response.created, response.output_text.delta,
response.completed, etc.)
- Full SSE event support with both event: and data: lines
- Configuration via gateway.http.endpoints.responses.enabled
The endpoint is disabled by default and can be enabled independently
from the existing Chat Completions endpoint.
Phase 1 implementation supports:
- String or ItemParam[] input
- system/developer/user/assistant message roles
- function_call_output items
- instructions parameter
- Agent routing via headers or model parameter
- Session key management
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
When resolving models from custom provider configurations, ensure the
provider name is attached to each inline model entry. This fixes model
resolution for custom providers where the model definition exists in
the config but lacks an explicit provider field.
Without this fix, inline models from custom providers (like amazon-bedrock)
would fail to resolve because the provider context was lost during the
flatMap operation.
Add AWS Bedrock Converse Stream API to the list of supported model APIs,
enabling custom provider configurations for Amazon Bedrock endpoints.
This allows users to configure Bedrock models in their clawdbot.json:
"models": {
"providers": {
"amazon-bedrock": {
"baseUrl": "https://bedrock-runtime.us-east-1.amazonaws.com",
"api": "bedrock-converse-stream",
"models": [...]
}
}
}
The underlying adapter already exists; this change exposes it as a valid
configuration option.
- Keep the selected chat session in ?session=... for deep links and reloads.\n- Only apply the query param on the Chat tab (avoid leaking it across navigation).\n- Render session <option> entries with stable keys to prevent label glitches.
Chat should scroll inside the thread, not the whole page.\n\n- Constrain the app shell to the viewport and disable outer scrolling.\n- Hide page-level scrolling for the chat tab so only .chat-thread scrolls.
Use SearchableSelectList for /sessions and /agents pickers,
matching the /models picker behavior.
- Session picker: search by session key, display name, or date
- Agent picker: search by agent ID or name
🤖 AI-assisted (Claude)
Tests now verify structure and content preservation rather than
checking for specific ANSI escape codes, which may not be present
in CI environments without TTY/color support.
* Web: trim HTML error bodies in web_fetch
* fix: trim web_fetch HTML error bodies (#1193) (thanks @sebslight)
---------
Co-authored-by: Sebastian Slight <sbarrios93@gmail.com>
Co-authored-by: Peter Steinberger <steipete@gmail.com>
- Add SearchableSelectList component with fuzzy filtering
- Integrate with /models command for quick model search
- Support up/down navigation while typing
- Uses pi-tui's fuzzyFilter for intelligent matching
Add shared parseBooleanValue()/isTruthyEnvValue() and apply across CLI, gateway, memory, and live-test flags for consistent env handling.
Introduce route-first fast paths, lazy subcommand registration, and deferred plugin loading to reduce CLI startup overhead.
Centralize config validation via ensureConfigReady() and add config caching/deferred shell env fallback for fewer IO passes.
Harden logger initialization/imports and add focused tests for argv, boolean parsing, frontmatter, and CLI subcommands.
When context limit is exceeded, the error message now suggests
setting agents.defaults.compaction.reserveTokensFloor to 4000
or higher to prevent future occurrences.
The openrouter/auto model is OpenRouter's internal routing feature,
not a callable model. While it's valid as a default (set automatically
during OpenRouter auth flow), showing it in the configure wizard's
model selection causes "Unknown model: openrouter/auto" errors when
users select it manually.
Add a HIDDEN_ROUTER_MODELS set to filter out such internal models
from the selection list while preserving existing functionality.
Fixes#1115
Co-Authored-By: Claude <noreply@anthropic.com>
Previously, when auto-compaction failed due to context overflow, the system
would reset the session and silently continue the execution loop without
sending any response to the user. This made it appear as if messages were
being ignored.
This change ensures users receive a clear error message explaining that
the context limit was exceeded and the conversation has been reset,
consistent with how role ordering conflicts are already handled.
Fixes the silent failure case where message + compaction exceeds context limits.
- Switch from inline to URL-based TwiML for outbound calls
- Store TwiML content temporarily and serve on webhook request
- Add twimlStorage map and cleanup helper methods
- Fix TwiML serving to handle CallStatus='in-progress' on initial request
Closes#864
- Enhanced error handling in image reference detection to skip malformed file URLs without crashing.
- Updated media loading logic to throw an error for invalid file URLs, ensuring better feedback for users.
- Added support for detecting file URLs in prompts using fileURLToPath for accurate path resolution.
- Updated image loading logic to default to JPEG format for optimized image processing.
- Improved error handling in image optimization to continue processing on failures.
- Added handling for file URLs using fileURLToPath for proper resolution.
- Updated logic to skip relative path resolution if ref.resolved is already absolute.
- Enhanced cap calculation for image loading to handle undefined maxBytes more gracefully.
- Auto-detect and load images referenced in user prompts
- Inject history images at their original message positions
- Fix EXIF orientation - rotate before resizing in resizeToJpeg
- Sandbox security: validate paths, block remote URLs when sandbox enabled
- Prevent duplicate history image injection across turns
- Handle string-based user message content (convert to array)
- Add bounds check for message index in history processing
- Fix regex to properly match relative paths (./ ../)
- Add multi-image support for iMessage attachments
- Pass MAX_IMAGE_BYTES limit to image loading
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Added afterEach hooks with server/ws cleanup to:
- server.channels.test.ts (3 tests)
- server.config-apply.test.ts (2 tests)
- server.sessions-send.test.ts (already had this)
This ensures ports are properly released between tests, preventing
timeout issues from port conflicts.
When a prompt submission fails with quota or rate limit errors, throw
FailoverError instead of the raw promptError. This enables the model
fallback system to try alternative models.
Previously, rate limit errors during the prompt phase (before streaming)
were thrown directly, bypassing fallback. Only response-phase errors
triggered model fallback.
Now checks if fallback models are configured and the error is failover-
eligible. If so, wraps in FailoverError to trigger the fallback chain.
- CLI for reading only (Twitter flags CLI posts as automated)
- Browser tool with paste hack for writing
- React input workaround with ClipboardEvent
- Selectors and rate limiting tips
- Credit: Shadow's Ruby documented the forbidden arts
The frontmatter parser was using a simple line-by-line regex that only
captured single-line key-value pairs. This meant multi-line metadata
blocks (as used by bundled hooks) were not parsed correctly.
Changes:
- Add extractMultiLineValue() to handle indented continuation lines
- Use JSON5 instead of JSON.parse() to support trailing commas
- Add comprehensive test coverage for frontmatter parsing
Fixes#1113
Strip prerelease (`-beta.1`) and build (`-4`) suffixes from the patch
component before parsing as integer. Previously `2026.1.11-4` parsed to
`patch: 0` because `Int("11-4")` returns nil; now correctly yields
`patch: 11`.
Fixes#1107
Co-Authored-By: Claude <noreply@anthropic.com>
When using inlineButtons="dm" or "group" scope, the validation check
resolveTelegramTargetChatType() failed for numeric chat IDs because
normalizeTelegramMessagingTarget() adds a "telegram:" prefix during
target resolution.
For example, target "5232990709" becomes "telegram:5232990709" after
normalization, but the regex /^-?\d+$/ expects a pure numeric string.
The fix strips the telegram: prefix before checking the numeric pattern.
Adds tests for resolveTelegramTargetChatType with various input formats.
On macOS, watching deep dependency trees can exhaust file descriptors and lead to spawn EBADF failures. The skills watcher only needs to observe skill changes, so ignore dotfiles, node_modules, and dist by default. Adds regression coverage.
When capabilities is configured as an object (e.g., { inlineButtons: "dm" })
instead of a string array, normalizeCapabilities() would crash with
"capabilities.map is not a function".
This can occur when using the new Telegram inline buttons scoping feature:
channels.telegram.capabilities.inlineButtons = "dm"
The fix adds an Array.isArray() guard to return undefined for non-array
capabilities, allowing channel-specific handlers (like
resolveTelegramInlineButtonsScope) to process the object format separately.
Fixes crash when using object-format TelegramCapabilitiesConfig.
For remote CDP profiles (e.g., Browserless), tab operations now use Playwright's
persistent connection instead of stateless HTTP requests. This ensures tabs
persist across operations rather than being terminated after each request.
Changes:
- pw-session.ts: Add listPagesViaPlaywright, createPageViaPlaywright, and
closePageByTargetIdViaPlaywright functions using the cached Playwright connection
- pw-ai.ts: Export new functions for dynamic import
- server-context.ts: For remote profiles (!cdpIsLoopback), use Playwright-based
tab operations; local profiles continue using HTTP endpoints
- server-context.ts: Fix ensureTabAvailable to not require wsUrl for remote
profiles since Playwright accesses pages directly
This is a follow-up to #895 which added authentication support for remote CDP
profiles. The original PR description mentioned switching to persistent Playwright
connections for tab operations, but only the auth changes were merged.
- Treat message tool `channel` as provider hint for dedupe/suppression.
- Prefer NO_REPLY after message tool sends to avoid duplicate replies.
Co-authored-by: Sash Catanzarite <1166151+thesash@users.noreply.github.com>
Fixes#1030
The describeAccount function for WhatsApp was not returning the
linked field, causing the Channels status section to show
"Not linked" even when WhatsApp was properly linked and working.
The fix adds the linked field to describeAccount, set to the same
value as configured (Boolean(account.authDir)). This ensures that
the Channels section and Health section show consistent status.
Added three new test cases to verify the new country, search_lang, and ui_lang
parameters are correctly passed to the Brave Search API.
Co-Authored-By: Claude <noreply@anthropic.com>
- Add country, search_lang, and ui_lang optional parameters to runWebSearch
- Pass new parameters to Brave Search API as URL query parameters
- Update cache key to include localization parameters
- Wire parameters through createWebSearchTool execute function
Support ${VAR_NAME} syntax in any config string value, substituted at
config load time. Useful for referencing API keys and secrets from
environment variables without hardcoding them in the config file.
- Only uppercase env vars matched: [A-Z_][A-Z0-9_]*
- Missing/empty env vars throw MissingEnvVarError with path context
- Escape with $${VAR} to output literal ${VAR}
- Works with $include (included files also get substitution)
Closes#1009
- @NessZerra Linear CLI - manage Linear issues from terminal, works with Claude Code/Clawdbot
- @jules Beeper CLI - read/send/archive messages via Beeper Desktop MCP API
* fix: truncate skill command descriptions to 100 chars for Discord
Discord slash commands have a 100 character limit for descriptions.
Skill descriptions were not being truncated, causing command registration
to fail with an empty error from the Discord API.
* style: format
* style: format
When a subagent is spawned, it creates a new agent directory but has no
auth-profiles.json. This adds a fallback in ensureAuthProfileStore() to
inherit auth-profiles from the main agent when the subagent has none,
ensuring subagents can use OAuth tokens without manual file copying.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Rename `format` to `snapshotFormat` in browser-tool schema and implementation
- Rename `format` to `outputFormat` in canvas-tool schema and implementation
- The parameter name `format` conflicts with JSON Schema keyword, causing
Google Cloud Code Assist to reject the schema as invalid
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
* chore: ignore local identity files (IDENTITY.md, USER.md)
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
* chore: ignore local identity files (#1001) (thanks @gerardward2007)
* chore: format session status tool
* chore: format bash exec background abort test
---------
Co-authored-by: Claude Opus 4.5 <noreply@anthropic.com>
Co-authored-by: Peter Steinberger <steipete@gmail.com>
When exec returns early in background mode, the tool-call AbortSignal can fire and previously caused killProcessTree(SIGKILL). Ignore abort after yielding/backgrounding so background sessions keep running.
When --output-format is specified for GPT models, save files with
the correct extension (.webp, .jpeg, or .png) instead of always
using .png.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Auto-detect model and apply appropriate defaults for size/quality
- Add --background, --output-format, and --style parameters
- Enforce dall-e-3 count=1 limitation with automatic adjustment
- Omit quality parameter for dall-e-2 (not supported)
- Document model-specific parameters and supported values
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
* Agent: exclude google-antigravity from history downgrade hack
* Lint: fix formatting in test
* Lint: formatting and unused vars in test
* fix: preserve google-antigravity tool calls (#894) (thanks @mukhtharcm)
---------
Co-authored-by: Peter Steinberger <steipete@gmail.com>
Previously, when block streaming was disabled (the default), text generated
between tool calls would only appear after all tools completed. This was
because onBlockReply wasn't passed to the subscription when block streaming
was off, so flushBlockReplyBuffer() before tool execution did nothing.
Now onBlockReply is always passed, and when block streaming is disabled,
block replies are sent directly during tool flush. Directly sent payloads
are tracked to avoid duplicates in final payloads.
Also fixes a race condition where tool summaries could be emitted before
the typing indicator started by awaiting onAgentEvent in tool handlers.
Combine Multi-Agent Swarm card with new technical write-ups:
- Link to orchestrated-ai-articles repo (manifesto + architecture)
- Keep clawdspace link for agent sandboxing
- Add blog post link
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Add cleanupSuspendedCliProcesses() to kill accumulated suspended processes
from isolated sessions that don't share sessionIds (e.g., cron jobs).
- Only targets Clawdbot processes (--session-id pattern)
- Only kills suspended processes (state T)
- Only triggers when >10 processes accumulated
- Does not affect user's Claude Code sessions
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Add userToken and userTokenReadOnly (default: true) config fields
- Implement token routing: reads prefer user token, writes use bot token
- Add tests for token routing logic
- Update documentation with required OAuth scopes
User tokens enable reading DMs and private channels without requiring
bot membership. The userTokenReadOnly flag (true by default) ensures
the user token can only be used for reads, preventing accidental
sends as the user.
Required user token scopes:
- channels:history, channels:read
- groups:history, groups:read
- im:history, im:read
- mpim:history, mpim:read
- users:read, reactions:read, pins:read, emoji:read, search:read
Telegram General topic (id=1) has inconsistent API behavior:
- sendMessage: rejects explicit message_thread_id=1
- sendChatAction: requires message_thread_id=1 for typing to show
Split into two helper functions:
- buildTelegramThreadParams: excludes General topic for messages
- buildTypingThreadParams: includes General topic for typing
* feat(whatsapp): add debounceMs for batching rapid messages
Add a `debounceMs` configuration option to WhatsApp channel settings
that batches rapid consecutive messages from the same sender into a
single response. This prevents triggering separate agent runs for
each message when a user sends multiple short messages in quick
succession (e.g., "Hey!", "how are you?", "I was wondering...").
Changes:
- Add `debounceMs` config to WhatsAppConfig and WhatsAppAccountConfig
- Implement message buffering in `monitorWebInbox` with:
- Map-based buffer keyed by sender (DM) or chat ID (groups)
- Debounce timer that resets on each new message
- Message combination with newline separator
- Single message optimization (no modification if only one message)
- Wire `debounceMs` through account resolution and monitor tuning
- Add UI hints and schema documentation
Usage example:
{
"channels": {
"whatsapp": {
"debounceMs": 5000 // 5 second window
}
}
}
Default behavior: `debounceMs: 0` (disabled by default)
Verified: All existing tests pass (3204 tests), TypeScript compilation
succeeds with no errors.
Implemented with assistance from AI coding tools.
Closes#967
* chore: wip inbound debounce
* fix: debounce inbound messages across channels (#971) (thanks @juanpablodlc)
---------
Co-authored-by: Peter Steinberger <steipete@gmail.com>
* fix(model-picker): list each provider/model combo separately
Previously, /model grouped models by name and showed all providers
that offer the same model (e.g. 'claude-sonnet-4-5 — anthropic, google-antigravity').
This was confusing because:
1. Users couldn't tell which provider would be used when selecting by number
2. The display implied choice between providers but selection was automatic
Now each provider/model combination is listed separately so users
can explicitly select the exact provider they want.
- Remove model grouping in buildModelPickerItems
- Display format changed from 'model — providers' to 'provider/model'
- pickProviderForModel now returns the single provider directly
- Updated tests to reflect new behavior
* fix: simplify model picker entries (#970) (thanks @mcinteerj)
---------
Co-authored-by: Keith the Silly Goose <keith@42bolton.macnet.nz>
Co-authored-by: Peter Steinberger <steipete@gmail.com>
Add option to disable automatic read receipts for WhatsApp messages.
When set to false, Clawdbot will not mark messages as read (blue ticks).
Closes#344
Changes:
- Add sendReadReceipts to WhatsAppConfig and WhatsAppAccountConfig types
- Add sendReadReceipts to zod schemas for validation
- Add sendReadReceipts to ResolvedWhatsAppAccount with fallback chain
- Pass sendReadReceipts through to monitorWebInbox
- Gate sock.readMessages() call based on config option
Default behavior (true) is preserved - only explicitly setting false
will disable read receipts.
When replaying conversation history to Gemini, tool calls without
thought_signature are downgraded to text blocks like [Tool Call: ...].
This leaked internal technical info into user-facing chat messages.
Added stripDowngradedToolCallText filter alongside existing Minimax
filter to remove these text representations before extraction.
* docs: clarify subagent announce status
* Make subagent announce structured and include run outcome
* fix: stabilize sub-agent announce status (#835) (thanks @roshanasingh4)
---------
Co-authored-by: Peter Steinberger <steipete@gmail.com>
Adds support for template variables in `messages.responsePrefix` that
resolve dynamically at runtime with the actual model used (including
after fallback).
Supported variables (case-insensitive):
- {model} - short model name (e.g., "claude-opus-4-5", "gpt-4o")
- {modelFull} - full model identifier (e.g., "anthropic/claude-opus-4-5")
- {provider} - provider name (e.g., "anthropic", "openai")
- {thinkingLevel} or {think} - thinking level ("high", "low", "off")
- {identity.name} or {identityName} - agent identity name
Example: "[{model} | think:{thinkingLevel}]" → "[claude-opus-4-5 | think:high]"
Variables show the actual model used after fallback, not the intended
model. Unresolved variables remain as literal text.
Implementation:
- New module: src/auto-reply/reply/response-prefix-template.ts
- Template interpolation in normalize-reply.ts via context provider
- onModelSelected callback in agent-runner-execution.ts
- Updated all 6 provider message handlers (web, signal, discord,
telegram, slack, imessage)
- 27 unit tests covering all variables and edge cases
- Documentation in docs/gateway/configuration.md and JSDoc
Fixes#923
When a launchd service is uninstalled via bootout, macOS marks it as
disabled in /var/db/com.apple.xpc.launchd/disabled.*.plist. This persists
across reboots and plist recreation. Future bootstrap calls fail with
error 5 (I/O error) because the service is still marked disabled.
Fix: Call 'launchctl enable' before 'launchctl bootstrap' to clear the
disabled state, matching the fix already applied to the macOS Swift app
in GatewayLaunchAgentManager.swift.
Related: #306
The `channels.slack.requireMention` setting was defined in the schema
but never passed to `resolveSlackChannelConfig()`, which always
defaulted to `true`. This meant setting `requireMention: false` at the
top level had no effect—channels still required mentions.
Pass `slackCfg.requireMention` as `defaultRequireMention` to the
resolver and use it as the fallback instead of hardcoded `true`.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
When a Telegram group is upgraded to a supergroup, the chat ID changes
(e.g., -123456 → -100123456). This causes the bot to lose its group
configuration since it's keyed by chat ID.
This change:
- Adds handler for `message:migrate_to_chat_id` event
- Logs the migration (old_id → new_id) for visibility
- If the old chat ID has config in channels.telegram.groups, automatically:
- Copies the config to the new chat ID
- Removes the old chat ID entry
- Saves the updated config file
This eliminates the need to manually update clawdbot.json when groups
migrate to supergroups.
Add ability to delete messages in Telegram chats via the message tool.
Changes:
- Add deleteMessageTelegram function in send.ts
- Add deleteMessage action handler in telegram-actions.ts
- Add delete action support in telegram message plugin adapter
- Add deleteMessage to TelegramActionConfig type
- Update message tool description to mention delete action
Usage:
- Via message tool: action="delete", chatId, messageId
- Can be disabled via channels.telegram.actions.deleteMessage=false
Limitations (Telegram API):
- Bot can delete its own messages in any chat
- Bot can delete others' messages only if admin with "Delete Messages"
- Messages older than 48h in groups may fail to delete
Build reply/session context once (no post-hoc ctx mutation) and type into the actual delivery target.
Thanks @davidguttman.
Co-authored-by: David Guttman <david@davidguttman.com>
When Anthropic's API returns an overloaded_error (temporary capacity issue),
the raw JSON error was being sent to users instead of a friendly message.
Changes:
- Add overloaded error pattern detection
- Return user-friendly message for overloaded errors
- Classify overloaded as failover-worthy (triggers retry logic)
This PR includes three main improvements:
1. Tailscale Binary Detection with Fallback Strategies
- Added findTailscaleBinary() with multi-strategy detection:
* PATH lookup via 'which' command
* Known macOS app path (/Applications/Tailscale.app/Contents/MacOS/Tailscale)
* find /Applications for Tailscale.app
* locate database lookup
- Added getTailscaleBinary() with caching
- Updated all Tailscale operations to use detected binary
- Added TUI warning when Tailscale binary not found for serve/funnel modes
2. Custom Gateway IP Binding with Fallback
- New bind mode "custom" allowing user-specified IP with fallback to 0.0.0.0
- Removed "tailnet" mode (folded into "auto")
- All modes now support graceful fallback: custom (if fail → 0.0.0.0), loopback (127.0.0.1 → 0.0.0.0), auto (tailnet → 0.0.0.0), lan (0.0.0.0)
- Added customBindHost config option for custom bind mode
- Added canBindTo() helper to test IP availability before binding
- Updated configure and onboarding wizards with new bind mode options
3. Health Probe Password Auth Fix
- Gateway probe now tries both new and old passwords
- Fixes issue where password change fails health check if gateway hasn't restarted yet
- Uses nextConfig password first, falls back to baseConfig password if needed
Files changed:
- src/infra/tailscale.ts: Binary detection + caching
- src/gateway/net.ts: IP binding with fallback logic
- src/config/types.ts: BridgeBindMode type + customBindHost field
- src/commands/configure.ts: Health probe dual-password try + Tailscale detection warning + bind mode UI
- src/wizard/onboarding.ts: Tailscale detection warning + bind mode UI
- src/gateway/server.ts: Use new resolveGatewayBindHost
- src/gateway/call.ts: Updated preferTailnet logic (removed "tailnet" mode)
- src/commands/onboard-types.ts: Updated GatewayBind type
- src/commands/onboard-helpers.ts: resolveControlUiLinks updated
- src/cli/*.ts: Updated bind mode casts
- src/gateway/call.test.ts: Removed "tailnet" mode test
Claude's extended thinking feature generates thought_signature fields
(message IDs like "msg_abc123...") in content blocks. When these are
sent to Google's Gemini API, it expects Base64-encoded bytes and
rejects Claude's format with a 400 error.
This commit adds stripThoughtSignatures() to remove these fields from
assistant message content blocks during sanitization, enabling session
histories to be shared across different providers (e.g., Claude → Gemini).
Fixes cross-provider session bug where switching from Claude-thinking
to Gemini (or vice versa) would fail with:
"Invalid value at 'thought_signature' (TYPE_BYTES), Base64 decoding failed"
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Refactors tests to use a shared config object and adds stricter assertions to verify that the config is passed correctly to createTelegramBot. Ensures the bindings property is checked in the test expectations.
The Telegram webhook and monitor now accept and pass through accountId and config parameters, enabling routing and configuration per Telegram account. Tests have been updated to verify correct bot instantiation and DM routing based on accountId bindings.
When readConfigFileSnapshot encounters validation errors, it now:
1. Returns the resolved config data instead of empty object
2. Uses passthrough() on main schema to preserve unknown fields
This prevents config loss when:
- User has custom/unknown fields
- Legacy config issues are detected but config is otherwise valid
- Zod schema does not recognize newer fields
Fixes config being overwritten with empty object on validation failure.
When multiple Telegram accounts are configured, the Connections UI now
displays individual status cards for each account showing:
- Bot username and account ID
- Running/configured status
- Last inbound message time
- Per-account errors
Falls back to the existing summary view for single-account configs.
No changes needed to types (already added upstream).
Add validateAnthropicTurns() to merge consecutive user messages that can
occur when steering messages are injected during streaming. This prevents
the API from rejecting requests due to improper role alternation.
Changes:
- Add validateAnthropicTurns() function in pi-embedded-helpers.ts
- Integrate validation into sanitization pipeline in pi-embedded-runner.ts
- Add user-friendly error message for role ordering errors
- Add comprehensive tests for the new validation function
In forum supergroups, messages from the General topic arrive without
message_thread_id in updates, but sendChatAction requires one to display
the typing indicator in the correct topic.
Use message_thread_id=1 (Telegram's internal ID for General topic) as
fallback when messageThreadId is undefined in a forum chat.
- Enforces strict buffering when enforceFinalTag is enabled.
- Prevents 'thinking out loud' planning steps (e.g. '*Locating Manulife*') from leaking to WhatsApp.
- Hardens <final> tag stripping to remove nested/hallucinated tags.
- Added src/utils/provider-utils.ts to track reasoning provider logic
- Updated isReasoningTagProvider to loosely match 'google-antigravity' (fixes sub-models)
- Enabled enforceFinalTag in reply.ts when provider matches
- Verified <final> tag stripping logic in pi-embedded-subscribe.ts
- Updated pi-embedded-runner to use consistent provider check for prompt hints
On screens under 1100px, the nav groups were displaying in a confusing
grid-like pattern. This flattens all nav items into a single horizontal
scrollable row using display:contents to unwrap the group containers.
Also fixes the issue where collapsed nav groups would hide items on
mobile (where the toggle button is hidden), making them inaccessible.
The nullish coalescing operator (??) only skips null/undefined, not
empty strings. For direct WhatsApp messages, ctx.SenderId was an empty
string, causing senderRaw to be "" instead of falling through to the
valid ctx.SenderE164 value.
This caused commands like /status to be rejected with "unauthorized
sender" for self-chat WhatsApp messages.
Tested: Verified /status command now works correctly for self-chat
WhatsApp messages after the fix.
When block streaming is enabled with verbose=off, tool blocks are hidden
but their boundary information was lost. Text segments before and after
tool execution would get coalesced into a single message because the
coalescer had no signal that a tool had executed between them.
This adds an onBlockReplyFlush callback that fires on tool_execution_start,
allowing the block reply pipeline to flush pending text before the tool
runs. This preserves natural message boundaries even when tools are hidden.
Fixes the issue where:
text → [hidden tool] → text → rendered as one merged message
Now correctly renders as:
text → [hidden tool] → text → two separate messages
Co-diagnosed-by: Krill (Discord assistant)
Windows system utilities like ipconfig, systeminfo, etc. write directly to
the console via WriteConsole API instead of stdout. When Node.js spawns
cmd.exe with piped stdio, these utilities produce empty output.
Changes:
- Switch from cmd.exe to PowerShell on Windows (properly redirects output)
- Disable detached mode on Windows (PowerShell doesn't pipe stdout when detached)
- Add windowsHide option to prevent console window flashing
- Update tests to use PowerShell-compatible syntax (Start-Sleep, semicolons)
Ensure 'main' alias is always stored as 'agent:main:main' to prevent
duplicate entries. Also update loadSessionEntry to check both forms
when looking up entries.
Fixes duplicate main sessions in session store.
The previous prompt was too permissive about skipping announcements.
Updated to strongly encourage announcing results since the requester
is waiting for a response.
- Add 'You MUST announce your result' instruction
- Clarify ANNOUNCE_SKIP is only for complete failures
- Improve guidance on providing useful summaries
The previous immediate probe (timeoutMs: 0) only caught already-completed
runs. Cross-process spawns need to actually wait via agent.wait RPC for
the gateway to signal completion, then trigger the announce flow.
- Rename probeImmediateCompletion to waitForSubagentCompletion
- Use 10 minute wait timeout for agent.wait RPC
- Remove leftover debug console.log statements
- Replace free functions with IncludeProcessor class
- Simplify IncludeResolver interface: { readFile, parseJson }
- Break down loadFile into focused private methods
- Use reduce() for array include merging
- Cleaner separation of concerns
- Move $include resolution to src/config/includes.ts
- Simplify io.ts by importing from includes module
- Cleaner API: resolveConfigIncludes(obj, configPath, resolver?)
- Re-export errors from io.ts for backwards compatibility
- Rename test file to match module name
Adds support for splitting clawdbot.json into multiple files using the
$include directive. This enables:
- Single file includes: { "$include": "./agents.json5" }
- Multiple file merging: { "$include": ["./a.json5", "./b.json5"] }
- Nested includes (up to 10 levels deep)
- Sibling key merging with includes
Features:
- Relative paths resolved from including file
- Absolute paths supported
- Circular include detection
- Clear error messages with resolved paths
Use case: Per-client agent configs for isolated sandboxed environments
(e.g., legal case management with strict data separation).
- Use resolveExtraParams() which was defined but unused
- Create streamFn wrapper that injects config-driven params
- Apply to both compaction and run sessions
Config path: agents.defaults.models["provider/model"].params.temperature
Example:
agents.defaults.models["anthropic/claude-sonnet-4"].params.temperature = 0.7
agents.defaults.models["openai/gpt-4"].params.maxTokens = 8192
Tiny readme change that makes it less confusing.
The section title being "macOS app" makes it seem like the app is mandatory, when it is optional. Updated it to just "Apps"
Cross-agent subagent spawns wrote transcripts to the spawner's agent
directory instead of the target agent's directory. For example, when
main spawned a codex subagent with session key agent:codex:subagent:...,
the transcript went to agents/main/sessions/ instead of agents/codex/sessions/.
Pass sessionAgentId to resolveSessionFilePath so transcripts are written
to the correct agent's session directory.
Expose the existing model override capability via CLI flags:
- Add --model to cron add and cron edit commands
- Document model and thinking overrides in cron-jobs.md
- Add CLI example showing model/thinking usage
The backend already supported model in agentTurn payloads;
this change exposes it through the CLI interface.
When the update runner passes custom env vars (like CLAWDBOT_UPDATE_IN_PROGRESS),
the current code uses `env ?? process.env` which replaces the entire environment
instead of merging — losing PATH, HOME, etc.
This causes the doctor step to fail with 'node: No such file or directory'.
Fix: merge custom env with process.env instead of replacing it.
Ensure 'main' alias is always stored as 'agent:main:main' to prevent
duplicate entries. Also update loadSessionEntry to check both forms
when looking up entries.
Fixes duplicate main sessions in session store.
The previous prompt was too permissive about skipping announcements.
Updated to strongly encourage announcing results since the requester
is waiting for a response.
- Add 'You MUST announce your result' instruction
- Clarify ANNOUNCE_SKIP is only for complete failures
- Improve guidance on providing useful summaries
The previous immediate probe (timeoutMs: 0) only caught already-completed
runs. Cross-process spawns need to actually wait via agent.wait RPC for
the gateway to signal completion, then trigger the announce flow.
- Rename probeImmediateCompletion to waitForSubagentCompletion
- Use 10 minute wait timeout for agent.wait RPC
- Remove leftover debug console.log statements
Cross-agent subagent spawns wrote transcripts to the spawner's agent
directory instead of the target agent's directory. For example, when
main spawned a codex subagent with session key agent:codex:subagent:...,
the transcript went to agents/main/sessions/ instead of agents/codex/sessions/.
Pass sessionAgentId to resolveSessionFilePath so transcripts are written
to the correct agent's session directory.
- Move config from messages.ackReaction to whatsapp.ackReaction
- New structure: {emoji, direct, group} with granular control
- Support per-account overrides in whatsapp.accounts.*.ackReaction
- Add Zod schema validation for new config
- Maintain backward compatibility with old messages.ackReaction format
- Update tests to new config structure (14 tests, all passing)
- Add comprehensive documentation in docs/providers/whatsapp.md
- Timing: reactions sent immediately upon message receipt (before bot reply)
Breaking changes:
- Config moved from messages.ackReaction to whatsapp.ackReaction
- Scope values changed: 'all'/'direct'/'group-all'/'group-mentions'
→ direct: boolean + group: 'always'/'mentions'/'never'
- Old config still supported via fallback for smooth migration
- Fix bug where ack reaction was not sent when group activation is 'always'
- When requireMention=false (activation: always), always send reaction
- Add test case for activation='always' scenario
- Update inline comments for clarity
- Add automatic emoji reactions on inbound WhatsApp messages
- Support all ackReactionScope modes: all, direct, group-all, group-mentions
- Reaction is sent AFTER successful reply (unlike Telegram/Discord)
- Errors are logged with proper context
- Add comprehensive test suite for ack reaction logic
Config usage:
messages:
ackReaction: "👀"
ackReactionScope: "group-mentions" # default
Closes: WhatsApp ack-reaction feature request
WhatsApp group mentions using the new Linked ID format (@lid) were not
being detected because jidToE164() was called without the authDir needed
to find the LID reverse mapping files.
Now isBotMentioned() and debugMention() accept an optional authDir
parameter, which is passed through from account.authDir.
OpenAI/ChatGPT returns "You have hit your ChatGPT usage limit (plus plan)"
when users exceed their plan quota. This error wasn't being recognized as a
rate limit, so fallback to alternative models wasn't triggering.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- runDaemonRestart() now returns Promise<boolean> indicating success
- update command only shows success when restart actually happened
- Fixes missing reasoningLevel type in compactEmbeddedPiSession
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Carl Ulsøe Christensen <carlulsoe@users.noreply.github.com>
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Images were previously converted to markdown data URLs which Claude API
treats as plain text, not as actual images.
Changes:
- Add parseMessageWithAttachments() that returns {message, images[]}
- Pass images through the stack to session.prompt() as content blocks
- Filter null/empty attachments before parsing
- Strip data URL prefix if client sends it
This enables iOS and other clients to send images that Claude can actually see.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
* fix(signal): handle reactions inside dataMessage.reaction
Signal reactions can arrive in two formats:
1. envelope.reactionMessage (already handled)
2. envelope.dataMessage.reaction (now handled)
The signal-cli SSE events use the second format, which was being
misinterpreted as a message with attachments, leading to 'broken
media / attachments' errors.
Changes:
- Add reaction property to SignalDataMessage type
- Check both envelope.reactionMessage and dataMessage.reaction
- Improve body content detection to properly identify reaction-only messages
- Add test for dataMessage.reaction format
* fix(signal): reaction notifications work when account is phone number
When reactionNotifications mode is 'own', notifications would never fire
because resolveSignalReactionTarget() returned a UUID but
shouldEmitSignalReactionNotification() compared it against the account
phone number, which never matched.
The fix:
- Add optional 'phone' field to SignalReactionTarget type
- Extract phone number first in resolveSignalReactionTarget(), include
it even when UUID is present
- In shouldEmitSignalReactionNotification() 'own' mode, check phone
match first before falling back to UUID comparison
This ensures reactions to your own messages are properly detected when
the Signal account is configured as a phone number and the reaction
event contains both targetAuthor (phone) and targetAuthorUuid.
* fix(signal): include phone in reaction target for own-mode matching
When targetAuthorUuid is present, also store targetAuthor phone number
in the reaction target. This allows own-mode reaction notifications to
match when comparing account phone against UUID-based targets.
The default Gmail hook path configured by `clawdbot hooks gmail setup` is `/gmail-pubsub`. Tailscale strips the mount path before proxying, so the request lands on `/` and the hook 404s under the default configuration.
When Tailscale is enabled, always listen on `/` internally and keep the public URL on the configured path (defaulting to `/gmail-pubsub`). This makes default and custom paths work reliably.
Alternative (not implemented here): call tailscale with a full target URL so the backend keeps the path, e.g. `tailscale funnel --set-path /gmail-pubsub http://127.0.0.1:8788/gmail-pubsub`. We did not take this path because it requires changing the CLI invocation to pass URLs (not ports) plus extra validation, which is a larger behavior change.
Fixes#652
When cron jobs with sessionTarget:"main" have wakeMode:"now",
they were being marked as completed immediately without waiting for the
agent to actually process the system event.
The issue was that requestHeartbeatNow() is fire-and-forget and
doesn't wait for the heartbeat to complete. The job would finish
with durationMs: 0 before the agent had a chance to run.
This fix:
- Adds runHeartbeatOnce to CronServiceDeps
- Wires it up in gateway/server.ts to load config and pass runtime
- Modifies executeJob() to call runHeartbeatOnce when wakeMode:"now"
- Waits for heartbeat to complete and maps status to cron result:
* "ran" → "ok"
* "skipped" → "skipped"
* "failed" → "error"
- Falls back to old behavior for wakeMode:"next-heartbeat" or if
runHeartbeatOnce is not available (backward compatibility)
Benefits:
- Jobs now have accurate durationMs reflecting actual processing time
- Jobs are correctly marked with "error" status if heartbeat fails
- Prevents race condition where job completes before agent runs
[AI-assisted] - Generated with z.ai GLM-4.7
[Tested: Lightly tested - Logic validated with test scenarios, code quality checks passed, integration testing requires live Clawdbot instance]
- Replace model catalog with 11 current models: gpt-5.1-codex, claude-opus-4-5,
gemini-3-pro, alpha-glm-4.7, gpt-5.1-codex-mini, gpt-5.1, glm-4.7-free,
gemini-3-flash, gpt-5.1-codex-max, minimax-m2.1-free, gpt-5.2
- Add accurate per-token costs from OpenCode Zen pricing
- Add accurate context windows and output limits
- Update aliases for new model families (codex, glm, minimax)
- Remove deprecated models (sonnet, haiku, o-series, gemini-2.5)
- Expand schema scrubber to strip additional constraint keywords rejected
by Cloud Code Assist (examples, minLength, maxLength, minimum, maximum,
multipleOf, pattern, format, minItems, maxItems, uniqueItems,
minProperties, maxProperties)
- Extend tool call ID sanitization to cover toolUse and toolCall block
types (previously only functionCall was sanitized)
- Update pi-tools test to include 'examples' in unsupported keywords
Fixes 400 errors when using google-antigravity/claude-opus-4-5-thinking:
- tools.N.custom.input_schema: JSON schema is invalid
- messages.N.content.N.tool_use.id: String should match pattern
Both node-pairing.ts and voicewake.ts were using a local defaultBaseDir()
that hardcoded ~/.clawdbot, ignoring CLAWDBOT_STATE_DIR.
This caused nodes/paired.json and settings/voicewake.json to be stored
in ~/.clawdbot instead of the configured state directory.
Fixes the bug where paired nodes config was stored in a different
location than the rest of the gateway state.
Adds `agent.humanDelay` config option to create natural rhythm between
streamed message bubbles. When enabled, introduces a random delay
(default 800-2500ms) between block replies, making multi-message
responses feel more like natural human texting.
Config example:
```json
{
"agent": {
"blockStreamingDefault": "on",
"humanDelay": {
"enabled": true,
"minMs": 800,
"maxMs": 2500
}
}
}
```
- First message sends immediately
- Subsequent messages wait a random delay before sending
- Works with iMessage, Signal, and Discord providers
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Previously, Read/Write/Edit tools used the global tool instances from
pi-coding-agent which had process.cwd() baked in at import time. Since
the gateway starts from /root/dev/ai/clawdbot, relative paths like
'SOUL.md' would incorrectly resolve there instead of the agent's
workspace (/root/clawd).
This fix:
- Adds workspaceDir option to createClawdbotCodingTools
- Creates fresh Read/Write/Edit tools bound to workspaceDir
- Adds cwd option to Bash tool defaults for consistency
- Passes effectiveWorkspace from pi-embedded-runner
Absolute paths and ~/... paths are unaffected. Sandboxed sessions
continue to use sandbox root as before.
Includes tests for Read/Write/Edit workspace path resolution.
Applies the same Swift 6 compatibility patterns from PR #166 (macOS) to the iOS app.
Changes:
- LocationService.swift: Added Sendable constraint to withTimeout<T> generic,
made CLLocationManagerDelegate methods nonisolated with Task { @MainActor in }
pattern to safely access MainActor state
- TalkModeManager.swift: Fixed OSLog string interpolation to avoid operator
overload issues with OSLogMessage in Swift 6
Addresses #164
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
New community showcase entries from Discord #showcase:
- ParentPay School Meals (@George5562) - UK school meal automation
- R2 Upload (@julianengel) - presigned URL file sharing
- iOS App via Telegram (@coard) - full iOS app built via chat
- Oura Ring Health Assistant (@AS) - health/calendar integration
When Claude CLI credentials (anthropic:claude-cli) expire, automatically
refresh using the stored refresh token instead of failing with
"No credentials found" error.
Changes:
- Read refreshToken from Claude CLI and store as OAuth credential type
- Implement bidirectional sync: after refresh, write new tokens back to
Claude Code storage (file on Linux/Windows, Keychain on macOS)
- Prefer OAuth over Token credentials (enables auto-refresh capability)
- Maintain backward compatibility for credentials without refreshToken
This enables long-running agents to operate autonomously without manual
re-authentication when OAuth tokens expire.
Co-Authored-By: Claude <noreply@anthropic.com>
When reasoning is enabled on non‑block providers, we now ignore interim streaming chunks and send only the final assistant answer at completion, so replies aren’t partial or duplicated.
* docs(telegram): Add @userinfobot tip with screenshot and privacy note
* docs: adjust telegram userinfobot tip
---------
Co-authored-by: sheeek <gitlab@ott.team>
Co-authored-by: Ayaan Zaidi <zaidi@uplause.io>
- Keep audioAsVoice-only payloads from being filtered out
- Allow empty payloads through when they carry the flag
- Remove temporary debug logs around audioAsVoice buffering
Co-authored-by: Manuel Hettich <17690367+ManuelHettich@users.noreply.github.com>
- Add [[audio_as_voice]] detection to splitMediaFromOutput()
- Pass audioAsVoice through onBlockReply callback chain
- Buffer audio blocks during streaming, flush at end with correct flag
- Non-audio media still streams immediately
- Fix: emit payloads with audioAsVoice flag even if text is empty
Co-authored-by: Manuel Hettich <17690367+ManuelHettich@users.noreply.github.com>
The heartbeat prompt from agents.defaults.heartbeat.prompt was being
injected into the system prompt for ALL agents, causing non-default
agents to read the default agent's identity files and adopt its persona.
Now the heartbeat prompt is only included when the session's agent ID
matches the configured default agent. Other agents receive no heartbeat
section in their system prompt.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Fix disableBlockStreaming logic in telegram/bot.ts to properly enable
block streaming when telegram.streamMode is 'block' regardless of
blockStreamingDefault setting
- Set minChars default to 1 for Telegram block mode so chunks send
immediately on newlines/sentences instead of waiting for 800 chars
- Skip coalescing for Telegram block mode when not explicitly configured
to reduce chunk batching delays
- Fix newline preference to wait for actual newlines instead of breaking
on any whitespace when buffer is under maxChars
Fixes issue where all Telegram messages were batched into one message
at the end instead of streaming as separate messages during generation.
- Add sanitizeToolCallId() to fix Cloud Code Assist tool call ID validation
- Apply sanitization in sanitizeSessionMessagesImages() for toolResult and assistant messages
- Add legacy CONFIG_PATH_CLAWDIS and STATE_DIR_CLAWDIS exports for backward compatibility
- Resolves Cloud Code Assist rejection of invalid tool call IDs with pipe characters
- Fixes missing session export functions that were blocking system startup
Addresses Cloud Code Assist API 400 errors from invalid tool call IDs like 'call_abc123|item_456'
Error messages in sendMessageWhatsApp, sendReactionWhatsApp, and
sendPollWhatsApp included hardcoded \"WhatsApp\" references. This caused
confusion when cron jobs using other providers (e.g., Telegram) failed
with errors mentioning WhatsApp.
Changes error messages to be provider-agnostic while maintaining the
same error handling behavior.
Fixes#461Fixes#470
Co-Authored-By: Claude <noreply@anthropic.com>
The --smoke qr argument was removed from the relay CLI, but the smoke
test remained in the macOS app packaging script. This caused the build
to fail with exit code 1 when the bundled relay didn't recognize the
argument.
Removes the 2-line smoke test section to fix the build.
Fixes#317
Co-Authored-By: Claude <noreply@anthropic.com>
- Add SignalReactionMessage type with emoji, targetAuthor, timestamp
- Handle reaction messages in monitor (log and skip for now)
- Prevents reactions from showing as unknown media
- Update MessageToolOptions type to include Slack threading options
- Remove duplicate threadTs property in slack/actions.ts
- Fix replyThreadTs parameter name in monitor.ts
- Update test to correctly verify 'first' mode threading behavior:
- 'off' mode: no threading unless already in a thread
- 'first' mode: first reply starts a thread
- Add new test case for 'first' mode threading
- Add shared hasRepliedRef state between auto-reply and tool paths
- Extract buildSlackThreadingContext helper in agent-runner.ts
- Extract resolveThreadTsFromContext helper in slack-actions.ts
- Update docs with clear replyToMode table (off/first/all)
- Add tests for first mode behavior across multiple messages
Add 30-second timeout after WebSocket opens to detect when Discord
never sends HELLO (zombie state). If isConnected stays false after
timeout, forces a fresh connection instead of hanging indefinitely.
Relates to #595
Add listener for Carbon GatewayPlugin 'debug' events to capture WebSocket
connection state changes. Critical events (close, reconnect, resume) are
logged at INFO level always; all debug messages logged in verbose mode.
Fixes#595
Enables multiple agents to process the same message simultaneously,
allowing teams of specialized agents with atomic tasks to work together
in the same group using one phone number.
Key features:
- Configure multiple agents per WhatsApp group/DM via routing.broadcast
- Parallel (default) or sequential processing strategies
- Full session isolation (separate history, workspace, tools per agent)
- Minimal code changes (~50 lines in auto-reply.ts)
- Backward compatible with existing routing
Use cases:
- Specialized agent teams (code reviewer + security scanner + docs)
- Multi-language support (EN + DE + ES agents)
- Quality assurance workflows (support + QA agents)
- Task automation (tracker + logger + reporter)
Example config:
{
"routing": {
"broadcast": {
"strategy": "parallel",
"120363403215116621@g.us": ["alfred", "baerbel", "assistant3"]
}
}
}
This enables scaling to hundreds of focused micro-agents on a single
phone number, each handling specific atomic tasks.
## Problem
When messages arrived while the agent was busy processing a previous message,
the same message could be enqueued multiple times into the followup queue.
This happened because Discord's event system can emit the same message multiple
times (e.g., during reconnects or due to slow listener processing), and the
followup queue had no deduplication logic.
This caused the bot to respond to the same user message 2-4+ times.
## Solution
Add simple exact-match deduplication in `enqueueFollowupRun()`: if a prompt
is already in the queue, skip adding it again. Extracted into a small
`isPromptAlreadyQueued()` helper for clarity.
## Testing
- Added test cases for deduplication (same prompt rejected, different accepted)
- Manually verified on Discord: single response per message even when multiple
events fire during slow agent processing
Recent changes added recordProviderActivity calls with accountId, but
the type definition and usage didn't include accountId in ActiveWebSendOptions.
This fix adds the optional accountId field and uses optional chaining
when accessing it to handle cases where options is undefined.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Replaces Type.Integer with Type.Number and simplifies the sessions_send tool schema to avoid anyOf/oneOf unions that cause 400 errors with Google Cloud Code Assist API.
Add --auth-choice minimax-api for direct MiniMax API usage at
https://api.minimax.io/anthropic using the anthropic-messages API.
Changes:
- Add applyMinimaxApiConfig() function with provider/model config
- Add minimax-api to AuthChoice type and CLI options
- Add handler and non-interactive support
- Fix duplicate minimax entry in envMap
- Update live test to use anthropic-messages API
- Add 11 unit tests covering all edge cases
- Document configuration in gateway docs
Test results:
- 11/11 unit tests pass
- 1/1 live API test passes (verified with real API key)
Co-Authored-By: Claude <noreply@anthropic.com>
When Antigravity returns 429, throw immediately instead of waiting for the
server-provided retry delay (which can be 10+ minutes). This lets clawdbot
quickly rotate to another account.
This patch was previously in PR #454 but was accidentally removed in 0dbb569
when bumping to pi-ai 0.40.0. The upstream release did NOT include this fix.
Context: Antigravity rate limits cause pi-ai to sleep for the full retry
delay inside the request, blocking the thread. Clawdbot's timeout would
eventually fire, but waiting 10+ minutes is unacceptable UX.
Bundled with the empty error message filter since both handle 429 recovery.
When 429/500 errors occur during tool execution, empty assistant messages
with stopReason='error' and content=[] get persisted to the session file.
These break the tool_use -> tool_result chain that Claude/Gemini require:
- Claude expects every tool_use block to have a matching tool_result
- Empty error messages inserted mid-sequence violate this invariant
- Results in: 'tool_use ids were found without tool_result blocks'
This patch filters out empty error messages when building session context,
allowing sessions to recover gracefully from transient API errors.
Evidence from production:
- 113 of 170 sessions had empty error messages
- Session 30764430 demonstrated recovery: 429 at 14:30:11 IST,
resumed successfully at 14:30:22, completed at 14:30:34
Sorry Peter, one more patch! 🙈
When identity.name is configured and responsePrefix is not explicitly set,
automatically default responsePrefix to [identity.name].
This means users only need to set their identity once:
{ identity: { name: "MyBot" } }
And outbound messages will automatically be prefixed with [MyBot].
When identity.name is configured, use it for the default messagePrefix
instead of hardcoded '[clawdbot]'. Falls back to 'clawdbot' if not set.
This allows users to customize how their bot identifies itself in messages
by setting identity.name in their config or IDENTITY.md.
Changes the default base path from "/" to "./" so the control UI works
correctly when served under a custom basePath (e.g., /jbclawd/).
Previously, assets were referenced with absolute paths like /assets/...,
which failed when the UI was served under a subpath. With relative paths
(./assets/...), the browser resolves them relative to the HTML location,
making the UI work regardless of the configured basePath.
- Add `label` field to session entries and expose it in `sessions.list`
- Display label column in the web UI sessions table
- Support `label` parameter in `sessions_send` for lookup by label instead of sessionKey
- `sessions.patch`: Accept and store `label` field
- `sessions.list`: Return `label` in session entries
- `sessions_spawn`: Pass label through to registry and announce flow
- `sessions_send`: Accept optional `label` param, lookup session by label if sessionKey not provided
- `agent` method: Accept `label` and `spawnedBy` params (stored in session entry)
- Add `label` column to sessions table in web UI
- Changed session store writes to merge with existing entry (`{ ...existing, ...new }`)
to preserve fields like `label` that might be set separately
We attempted to implement label persistence "properly" by passing the label
through the `agent` call and storing it during session initialization. However,
the auto-reply flow has multiple write points that overwrite the session entry,
and making all of them merge-aware proved unreliable.
The working solution patches the label in the `finally` block of
`runSubagentAnnounceFlow`, after all other session writes complete.
This is a workaround but robust - the patch happens at the very end,
just before potential cleanup.
A future refactor could make session writes consistently merge-based,
which would allow the cleaner approach of setting label at spawn time.
```typescript
// Spawn with label
sessions_spawn({ task: "...", label: "my-worker" })
// Later, find by label
sessions_send({ label: "my-worker", message: "continue..." })
// Or use sessions_list to see labels
sessions_list() // includes label field in response
```
Cloud Code Assist API requires strict JSON Schema draft 2020-12 compliance
and rejects keywords like patternProperties, additionalProperties, $schema,
$id, $ref, $defs, and definitions.
This extends cleanSchemaForGemini to:
- Remove all unsupported keywords from tool schemas
- Add oneOf literal flattening (matching existing anyOf behavior)
- Add test to verify no unsupported keywords remain in tool schemas
Add comprehensive tests for sandbox-formatters.ts (20 tests):
- formatStatus: running/stopped with emojis
- formatSimpleStatus: running/stopped without emojis
- formatImageMatch: match/mismatch indicators
- formatAge: seconds, minutes, hours, days with edge cases
- countRunning: count items by running state
- countMismatches: count items by image mismatch
All pure functions now covered. 152 LOC tests added.
Total test count: 39 tests (19 integration + 20 unit)
Test coverage increased from ~66% to ~80%.
- Remove unused normalizeOptions function
- Fix line length violations (format long lines)
- Fix import order (alphabetical)
- Format function signatures for readability
All lint checks now passing.
Improvements:
- Extract help text examples into EXAMPLES constant object
- Add createRunner() helper to reduce try-catch boilerplate
- Add normalizeOptions() helper (prepared for future use)
- Fix command names from 'clawd' to 'clawdbot' throughout
- Update main sandbox command to show help by default
- Better organize code with clear section comments
Increases from 82 to 137 LOC but with much better organization
and reduced duplication in error handling.
Update sandbox.ts to import and use functions from:
- sandbox-display.ts for all UI output
- sandbox-formatters.ts for data formatting
Removes 141 LOC of display/formatting code from sandbox.ts,
reducing it from 351 to 210 LOC (-40%).
Core business logic now clearer and more focused.
Improvements:
- Extract validation into separate function
- Split display logic from business logic
- Create reusable container matcher for agent filtering
- Abstract status/image formatting into helpers
- Reduce code duplication between containers and browsers
- Extract container removal into generic function
- Add type safety with FilteredContainers type
- Improve readability with smaller, focused functions
Changes:
- validateRecreateOptions(): Validate mutual exclusivity
- fetchAndFilterContainers(): Fetch + filter in one place
- createAgentMatcher(): Reusable agent filter predicate
- displayContainers/Browsers(): Dedicated display functions
- displaySummary/RecreatePreview/Result(): Clear separation
- removeContainer(): Generic removal with error handling
- Format helpers: formatStatus, formatImageMatch, etc.
- Count helpers: countRunning, countMismatches
Result: 85 more lines but much better maintainability and testability.
Add 'clawd sandbox list' and 'clawd sandbox recreate' commands to manage
sandbox containers. This fixes the issue where containers continue using
old images/configs after updates.
Problem:
- When sandbox Docker images or configs are updated, existing containers
keep running with old settings
- Containers are only recreated after 24h inactivity (pruning)
- If agents are used regularly, old containers run indefinitely
Solution:
- 'clawd sandbox list': Show all containers with status, age, and image match
- 'clawd sandbox recreate': Force container removal (recreated on next use)
- Supports --all, --session, --agent, --browser filters
- Requires confirmation unless --force is used
Implementation:
- Added helper functions to sandbox.ts (list/remove containers)
- Created sandbox-cli.ts following existing CLI patterns
- Created commands/sandbox.ts with list and recreate logic
- Integrated into program.ts
Use case: After updating sandbox images or changing sandbox config,
run 'clawd sandbox recreate --all' to ensure fresh containers.
- Add replyStyle config at global, team, and channel levels
- "thread" replies to the original message (for Posts layout channels)
- "top-level" posts as a new message (for Threads layout channels)
- Default based on requireMention: false → top-level, true → thread
- DMs always use thread style (direct reply)
- Add sendMessageMSTeams for proactive messaging via CLI/gateway
- Wire msteams into outbound delivery, heartbeat targets, and gateway send
- Fix reply delivery to use SDK's getConversationReference() for proper
bot info, avoiding "Activity Recipient undefined" errors
- Use proactive messaging for replies to post as top-level messages
(not threaded) by omitting activityId from conversation reference
- Add lazy logger in send.ts to avoid test initialization issues
- Add teams/channels config structure to MSTeamsConfig
- Implement requireMention check in monitor.ts
- Resolution order: channel > team > global > default (true)
- Update zod schema for validation
- Document RSC permissions for receiving all messages without @mention
- Document Graph API Proxy pattern for historical message access
- Document private channel limitations
- Document team/channel ID format (use URL path, not groupId)
- Add msteams to config-reload.ts (ProviderKind, ReloadAction, rules)
- Add msteams to PairingProvider for pairing code support
- Create conversation-store.ts for storing ConversationReference
- Implement DM policy check (disabled/pairing/open/allowlist)
- Fix WasMentioned to check actual bot mentions via entities
- Fix server shutdown by using custom Express server with httpServer.close()
- Pass authConfig to CloudAdapter for outbound call authentication
- Improve error logging with JSON serialization
- Integrate dispatchReplyFromConfig() for full agent routing
- Add msteams to TextChunkProvider and OriginatingChannelType
- Add msteams case to route-reply (proactive not yet supported)
- Strip @mention HTML tags from Teams messages
- Fix session key to exclude messageid suffix
- Add typing indicator support
- Add proper logging for debugging
- Add Microsoft 365 Agents SDK packages (@microsoft/agents-hosting,
@microsoft/agents-hosting-express, @microsoft/agents-hosting-extensions-teams)
- Add MSTeamsConfig type and zod schema
- Create src/msteams/ provider with monitor, token, send, probe
- Wire provider into gateway (server-providers.ts, server.ts)
- Add msteams to all provider type unions (hooks, queue, cron, etc.)
- Update implementation guide with new SDK and progress
- Direct link to Azure Bot creation page
- Field-by-field table for Project details
- Pricing tier options
- Microsoft App ID settings (Single Tenant, Create new)
- Note about SDK version requirement
Initial research and implementation guide for adding msteams as a new
messaging provider. Includes:
- Provider structure patterns from existing implementations
- Gateway integration requirements
- Config types and validation schemas
- Onboarding flow patterns
- MS Teams Bot Framework SDK considerations
- Files to create/modify checklist
This is exploratory work - implementation plan to follow.
Anthropic blocks specific lowercase tool names (bash, read, write, edit)
when using OAuth tokens. This fix:
1. Renames blocked tools to capitalized versions (Bash, Read, Write, Edit)
in pi-tools.ts via renameBlockedToolsForOAuth()
2. Passes all tools as customTools in splitSdkTools() to bypass
pi-coding-agent's built-in tool filtering, which expects lowercase names
The capitalized names work with both OAuth tokens and regular API keys.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Session names in the Sessions table are now clickable links
- Clicking navigates to /chat?session=<key> with that session loaded
- Global sessions excluded (not real conversations)
- Added .session-link CSS styling (accent color, underline on hover)
- Chat page reads 'session' query param and cleans URL after applying
Adds a brief 'Messaging' section to the system prompt to guide agents on:
- Reply in session = auto-routes to source provider
- Cross-session = use sessions_send
- Never use bash/curl for provider messaging
This helps prevent agents from using shell workarounds for messaging
when Clawdbot already handles routing internally.
Add KillMode=process to generated systemd unit file. Without this,
podman's conmon processes (which monitor sandbox containers) block
shutdown since they're children of the gateway process.
This preserves the desired behavior of keeping containers alive
across restarts while preventing systemd from waiting indefinitely.
Adds a brief 'Messaging' section to the system prompt to guide agents on:
- Reply in session = auto-routes to source provider
- Cross-session = use sessions_send
- Never use bash/curl for provider messaging
This helps prevent agents from using shell workarounds for messaging
when Clawdbot already handles routing internally.
Changes webchat textarea behavior:
- Enter key now sends message (was Cmd/Ctrl+Enter)
- Shift+Enter allows line breaks
- Updated placeholder text to reflect new behavior
Fixes#411
Co-Authored-By: Claude <noreply@anthropic.com>
The status bar refresh failed because refreshSessionInfo() compared
currentSessionKey (e.g. 'main') against the canonical session keys
returned by the gateway (e.g. 'agent:default:main').
This fix uses parseAgentSessionKey to also match canonical keys by
their 'rest' component, allowing 'main' to match 'agent:default:main'.
Fixes: status bar showing stale values after setting changes
AI-assisted: yes (Claude)
Testing: lightly tested (build passes, logic verified)
Users without a visible phone number (like some Signal users)
were being silently dropped. Now falls back to UUID for sender ID.
- Add sourceUuid to SignalEnvelope type
- Use sourceUuid when sourceNumber is not available
- Only check against bot account when sourceNumber exists (avoid UUID comparison issues)
Replace app icons with new Clawdbot branding (lobster-in-phone-booth design) across iOS, Android, and macOS.
Changes:
- iOS: Updated all 14 icon sizes in AppIcon.appiconset (20px to 1024px)
- Android: Updated launcher icons for all density buckets (mdpi to xxxhdpi)
- macOS: Updated Icon.icon bundle and regenerated Clawdbot.icns
Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
When cron jobs specify a Telegram topic in the 'to' field like
'-1003420657307:topic:60', the delivery code now correctly parses
this into separate chatId and messageThreadId values.
Previously, the entire string was passed as chatId, causing messages
to fail delivery to forum topics.
When delivering cron job output to Telegram, the 'to' field now supports
specifying a topic (forum thread) ID in addition to the chat ID.
Supported formats:
- chatId (plain chat ID or @username)
- chatId:topicId (chat ID with numeric topic ID)
- chatId:topic:topicId (alternative format with explicit marker)
This enables cron jobs to deliver messages to specific forum topics
rather than always going to the main/general topic.
Adds parseTelegramTarget helper function with unit tests.
(cherry picked from commit 24a6595e81)
Replace app icons with new Clawdbot branding (lobster-in-phone-booth design) across iOS, Android, and macOS.
Changes:
- iOS: Updated all 14 icon sizes in AppIcon.appiconset (20px to 1024px)
- Android: Updated launcher icons for all density buckets (mdpi to xxxhdpi)
- macOS: Updated Icon.icon bundle and regenerated Clawdbot.icns
Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
Heartbeat Telegram delivery was failing when the bot token was
configured only via telegram.botToken in config (without TELEGRAM_BOT_TOKEN
environment variable).
Root cause: deliverOutboundPayloads was called without accountId parameter,
so sendMessageTelegram couldn't determine which account to use and couldn't
find the token from config.
Fix: Resolve default Telegram accountId when provider is "telegram" and pass
it to deliverOutboundPayloads. This follows the same pattern used elsewhere in
the codebase (e.g., cron uses resolveTelegramToken).
Changes:
- Added import for resolveDefaultTelegramAccountId
- Added accountId resolution for telegram provider
- Updated deliverOutboundPayloads call to include accountId
Fixes#318
Co-Authored-By: Claude <noreply@anthropic.com>
Fixes parameter mismatch between AI tool schema and internal validation.
The TypeBox schema now uses `jobId` for update/remove/run/runs actions,
matching what users expect based on the returned job objects.
Changes:
- Changed parameter from `id` to `jobId` in TypeBox schema for update/remove/run/runs
- Updated execute function to read `jobId` parameter
- Updated tests to use `jobId` in input parameters
The gateway protocol still uses `id` internally - the tool now maps
`jobId` from the AI to `id` for the gateway call.
Fixes#185
Co-Authored-By: Claude <noreply@anthropic.com>
Heartbeat Telegram delivery was failing when the bot token was
configured only via telegram.botToken in config (without TELEGRAM_BOT_TOKEN
environment variable).
Root cause: deliverOutboundPayloads was called without accountId parameter,
so sendMessageTelegram couldn't determine which account to use and couldn't
find the token from config.
Fix: Resolve default Telegram accountId when provider is "telegram" and pass
it to deliverOutboundPayloads. This follows the same pattern used elsewhere in
the codebase (e.g., cron uses resolveTelegramToken).
Changes:
- Added import for resolveDefaultTelegramAccountId
- Added accountId resolution for telegram provider
- Updated deliverOutboundPayloads call to include accountId
Fixes#318
Co-Authored-By: Claude <noreply@anthropic.com>
Fixes parameter mismatch between AI tool schema and internal validation.
The TypeBox schema now uses `jobId` for update/remove/run/runs actions,
matching what users expect based on the returned job objects.
Changes:
- Changed parameter from `id` to `jobId` in TypeBox schema for update/remove/run/runs
- Updated execute function to read `jobId` parameter
- Updated tests to use `jobId` in input parameters
The gateway protocol still uses `id` internally - the tool now maps
`jobId` from the AI to `id` for the gateway call.
Fixes#185
Co-Authored-By: Claude <noreply@anthropic.com>
The macOS app was crashing in two scenarios:
1. Bundle.module crash (fixes#213): When the first tool event arrived,
ToolDisplayRegistry tried to load config via ClawdbotKitResources.bundle,
which used Bundle.module directly. In packaged apps, Bundle.module
couldn't find the resource bundle at the expected path, causing a
fatal assertion failure after ~40-80 minutes of runtime.
2. dyld crash (fixes#417): Swift 6.2 requires libswiftCompatibilitySpan.dylib
but SwiftPM doesn't bundle it automatically, causing immediate crash on
launch with "Library not loaded" error.
Changes:
- ClawdbotKitResources.swift: Replace direct Bundle.module access with a
safe locator that checks multiple paths and falls back gracefully
- package-mac-app.sh: Copy ClawdbotKit_ClawdbotKit.bundle to Resources
- package-mac-app.sh: Copy libswiftCompatibilitySpan.dylib from Xcode
toolchain to Frameworks
Tested on macOS 26.2 with Swift 6.2 - app launches and runs without crashes.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
The /status command was showing 'anthropic/claude-sonnet-4-5' even when
using 'google-antigravity/claude-sonnet-4-5' because buildStatusMessage
received only the model name without the provider prefix.
When resolveConfiguredModelRef parsed a model string without a slash,
it fell back to DEFAULT_PROVIDER ('anthropic'), causing the misleading
display.
Fix: Pass the full 'provider/model' string to buildStatusMessage so
the provider is correctly extracted and displayed.
🪿 Investigated and submitted by Keith the Silly Goose
Add proper type assertion for ApiClientOptions["fetch"] to resolve
TypeScript compilation errors when passing fetch implementation to
grammY's Bot constructor. This follows the same pattern already used
in bot.ts.
Fixes#465
The macOS app was crashing in two scenarios:
1. Bundle.module crash (fixes#213): When the first tool event arrived,
ToolDisplayRegistry tried to load config via ClawdbotKitResources.bundle,
which used Bundle.module directly. In packaged apps, Bundle.module
couldn't find the resource bundle at the expected path, causing a
fatal assertion failure after ~40-80 minutes of runtime.
2. dyld crash (fixes#417): Swift 6.2 requires libswiftCompatibilitySpan.dylib
but SwiftPM doesn't bundle it automatically, causing immediate crash on
launch with "Library not loaded" error.
Changes:
- ClawdbotKitResources.swift: Replace direct Bundle.module access with a
safe locator that checks multiple paths and falls back gracefully
- package-mac-app.sh: Copy ClawdbotKit_ClawdbotKit.bundle to Resources
- package-mac-app.sh: Copy libswiftCompatibilitySpan.dylib from Xcode
toolchain to Frameworks
Tested on macOS 26.2 with Swift 6.2 - app launches and runs without crashes.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Add signalToolStart to TypingSignaler and call it from onAgentEvent
when tools begin executing. This keeps the typing indicator visible
during long-running tool operations.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
Port Antigravity payload enhancements from CLIProxyAPI v6.6.89:
- Add ANTIGRAVITY_SYSTEM_INSTRUCTION with agent identity/guidelines
- Inject systemInstruction with role 'user' for Antigravity requests
- Add requestType: 'agent' to wrapped request body
- Update userAgent to 'antigravity' for Antigravity requests
This fixes RESOURCE_EXHAUSTED (429) errors when using Antigravity.
Adapted from: https://github.com/NoeFabris/opencode-antigravity-auth/pull/137
Reference: https://github.com/router-for-me/CLIProxyAPI/commit/67985d8
Previously, startTypingLoop would return early if the typing timer was
already running, which meant the TTL would never get refreshed during
long tool executions. This caused the typing indicator to stop after
2 minutes even if tools were still running.
Now we refresh the TTL at the start of startTypingLoop, before the
early-return checks. This keeps typing alive during long operations.
- Default: sendAudio (file with metadata) - preserves old behavior
- Opt-in: [[audio_as_voice]] tag for voice bubble
This is non-breaking - existing integrations keep working.
Allow agents to specify audio mode via inline tag:
- Default: voice bubble (sendVoice)
- [[audio_as_file]]: audio file with metadata (sendAudio)
The tag is stripped from the final message text.
Example agent response:
Here's a podcast episode! [[audio_as_file]]
MEDIA:https://example.com/episode.mp3
- Add audioAsVoice option to ReplyPayload type
- Update bot.ts to use sendVoice by default for audio (voice bubble)
- When audioAsVoice is false, use sendAudio (file with metadata)
This allows agents to control voice vs file mode via ReplyPayload.
Use Telegram's sendVoice API for audio files by default, displaying them
as round playable voice bubbles instead of file attachments.
Changes:
- Add asVoice option to TelegramSendOpts (defaults to true)
- When asVoice is true (default): use api.sendVoice() for voice bubbles
- When asVoice is false: use api.sendAudio() for traditional audio files
This gives callers control: voice messages for TTS/quick responses,
audio files for music/podcasts with metadata display.
Add automatic thinking mode support for Z.AI GLM-4.x models:
- GLM-4.7: Preserved thinking (clear_thinking: false)
- GLM-4.5/4.6: Interleaved thinking (clear_thinking: true)
Uses Z.AI Cloud API format: thinking: { type: "enabled", clear_thinking: boolean }
Includes patches for pi-ai, pi-agent-core, and pi-coding-agent to pass
extraParams through the stream pipeline. User can override via config
or disable via --thinking off.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Convert flat bullet list to Mintlify CardGroup/Card components
- Add icons and visual hierarchy
- Include author attribution and tags
- Add 'Submit Your Project' section with Steps component
- Organize by category with emoji headers
Makes the showcase more visually appealing and easier to scan.
- Keep the chosen mic label visible when it disconnects and show a disconnected hint while falling back to system default.
- Avoid clearing the preferred mic on device changes so it auto-restores when available.
- Add audio input change and default-input logs in voice wake runtime/tester/meter to debug routing.
Why: Start Test is a local verification tool, but it was forwarding transcripts to the gateway/chat, which confused users and made tests look like real commands.
What: stop forwarding from VoiceWakeTester and clarify in docs that the tester does not send to the gateway.
Why: voice wake tests often delivered partial/final transcripts without reliable word timings, so trigger matching failed, timeouts overwrote detections, and test runs/mic capture kept running after UI changes.
What: add text-only/prefix fallback and silence-based detection in the test flow, stop/clean up any prior test, cancel timeout on detection/stop, and tear down meter/test when the Voice Wake tab is inactive. Runtime detection now falls back on final text-only matches when timing is missing. UI state now reflects finalizing and prevents hanging tests.
Add topic-specific session file isolation to fix root cause of Gemini turn validation errors.
Each Telegram topic now maintains its own conversation history file, eliminating race
conditions and message corruption during concurrent topic processing.
Changes:
1. Enhanced resolveSessionTranscriptPath() to support optional topicId parameter
- Topic ID (Telegram messageThreadId) now incorporated into session filename
- Format: sessionId.jsonl (direct chats) vs sessionId-topic-{topicId}.jsonl (topics)
- Backward compatible: topicId is optional
2. Updated reply.ts to pass MessageThreadId to session file resolution
- ctx.MessageThreadId now flows through to resolveSessionTranscriptPath()
- Automatically provides topic context for each incoming message
3. Automatic propagation through entire system
- sessionFile parameter automatically carries topic-specific path through:
- FollowupRun object (queued runs)
- runEmbeddedPiAgent() calls
- compactEmbeddedPiSession() calls
- SessionManager lifecycle (load, read, write operations)
Benefits:
✓ Complete elimination of shared .jsonl race conditions
✓ Each topic's conversation history independently cached
✓ SessionManager instances operate on isolated files
✓ No concurrent mutations of the same message history
✓ Maintains full Phase 1 turn validation as safety layer
Testing:
✓ Build succeeds with no TypeScript errors
✓ Backward compatible with non-topic sessions (direct messages)
✓ Topic ID properly extracted from Telegram messageThreadId
Expected impact:
- Gemini "function call turn" errors eliminated (root cause fixed)
- Message history corruption prevented across all topics
- Improved stability in multi-topic scenarios
- Each topic maintains independent conversation state
This completes the two-phase fix:
- Phase 1 (previous): Turn validation to suppress errors
- Phase 2 (current): Topic isolation to fix root cause
🤖 Generated with Claude Code
Co-Authored-By: Claude Haiku 4.5 <noreply@anthropic.com>
Add conversation turn validation to prevent "400 function call turn comes immediately
after a user turn or after a function response turn" errors when using Gemini models
in multi-topic/multi-channel Telegram conversations.
Changes:
1. Added validateGeminiTurns() function to detect and fix turn sequence violations
- Merges consecutive assistant messages into single message
- Preserves metadata (usage, stopReason, errorMessage) from later message
- Handles edge cases: empty arrays, single messages, tool results
2. Applied validation at two critical message points in pi-embedded-runner.ts:
- Compaction flow (lines 674-678): Before compact() call
- Normal agent run (lines 989-993): Before replaceMessages() call
3. Comprehensive test coverage with 8 test cases:
- Empty arrays and single messages
- Alternating user/assistant sequences (no change needed)
- Consecutive assistant message merging with metadata preservation
- Tool result message handling
- Real-world corrupted sequences with mixed content types
Testing:
✓ All 7 test cases pass (pi-embedded-helpers.test.ts)
✓ Full build succeeds with no TypeScript errors
✓ No breaking changes to existing functionality
This is Phase 1 of a two-phase fix:
- Phase 1 (completed): Turn validation to suppress Gemini errors
- Phase 2 (pending): Root cause analysis of why history gets corrupted with topic switching
🤖 Generated with Claude Code
Co-Authored-By: Claude Haiku 4.5 <noreply@anthropic.com>
Add in-memory TTL-based caching to reduce file I/O bottlenecks in message processing:
1. Session Store Cache (45s TTL)
- Cache entire sessions.json in memory between reads
- Invalidate on writes to ensure consistency
- Reduces disk I/O by ~70-80% for active conversations
- Controlled via CLAWDBOT_SESSION_CACHE_TTL_MS env var
2. SessionManager Pre-warming
- Pre-warm .jsonl conversation history files into OS page cache
- Brings SessionManager.open() from 10-50ms to 1-5ms
- Tracks recently accessed sessions to avoid redundant warming
3. Configuration Support
- Add SessionCacheConfig type with cache control options
- Enable/disable caching and set custom TTL values
4. Testing
- Comprehensive unit tests for cache functionality
- Test cache hits, TTL expiration, write invalidation
- Verify environment variable overrides
This fixes the slowness reported with multiple Telegram topics/channels.
Expected performance gains:
- Session store loads: 99% faster (1-5ms → 0.01ms)
- Overall message latency: 60-80% reduction for multi-topic workloads
- Memory overhead: < 1MB for typical deployments
- Disk I/O: 70-80% reduction in file reads
Rollback: Set CLAWDBOT_SESSION_CACHE_TTL_MS=0 to disable caching
🤖 Generated with Claude Code
Co-Authored-By: Claude Haiku 4.5 <noreply@anthropic.com>
Previously, grammY's default bot.start() processed updates sequentially,
blocking all Telegram messages while one was being handled. This made
maxConcurrent settings ineffective for Telegram.
Now uses @grammyjs/runner which processes updates concurrently, matching
the behavior of Discord (Promise.all) and WhatsApp (fire-and-forget).
Benefits:
- Ack reactions (👀) appear immediately, not after queue clears
- Multiple chats can be processed in parallel
- maxConcurrent setting now works correctly for Telegram
- Long-running tool calls no longer block other conversations
When the conversation context exceeds the model's limit, instead of
throwing an opaque error or returning raw JSON, we now:
1. Detect context overflow errors (413, request_too_large, etc.)
2. Return a user-friendly message explaining the issue
3. Suggest using /new or /reset to start fresh
This prevents the assistant from becoming completely unresponsive
when context grows too large (e.g., from many screenshots or long
tool outputs).
Addresses issue #394
Claude API on Vertex AI (Cloud Code Assist) rejects nested anyOf schemas
as invalid JSON Schema draft 2020-12. This change:
- Add tryFlattenLiteralAnyOf() to convert Type.Union([Type.Literal(...)])
patterns from anyOf with const values to flat enum arrays
- Update stringEnum helper in bash-tools to use Type.Unsafe with flat enum
- Flatten BrowserActSchema from discriminated union to single object
- Simplify TelegramToolSchema to use Type.String() for IDs
Fixes 400 errors when sending messages through WhatsApp/Telegram providers.
Add docs/multi-agent-sandbox-tools.md covering:
- Configuration examples (personal + restricted, work agents)
- Different sandbox modes per agent
- Tool restriction patterns (read-only, safe execution, communication-only)
- Configuration precedence rules
- Migration guide from single-agent setups
- Troubleshooting tips
Add PR_SUMMARY.md for upstream submission with:
- Feature overview and use cases
- Implementation details (49 LoC across 5 files)
- Test coverage (18 new tests, all existing tests pass)
- Backward compatibility confirmation
- Migration examples
---
Kudos to Eula, the beautiful and selfless family owl 🦉
This feature was developed to enable safe, restricted access
for family group chats while maintaining full access for
the personal assistant. Schuhu!
Add new section explaining:
- How to configure per-agent sandbox settings
- How to configure per-agent tool restrictions
- Benefits (security isolation, resource control, flexible policies)
- Link to detailed guide
Include example config showing personal assistant (no sandbox)
vs family bot (sandboxed with read-only tools).
Add 5 tests for agent-specific tool restrictions:
- Apply global tool policy when no agent-specific policy exists
- Apply agent-specific tool policy
- Allow different tool policies for different agents
- Combine global and agent-specific deny lists
- Work with sandbox tools filtering
All tests pass.
Add 6 tests for agent-specific sandbox configuration:
- Use global sandbox config when no agent-specific config exists
- Override with agent-specific sandbox mode 'off'
- Use agent-specific sandbox mode 'all'
- Use agent-specific scope
- Use agent-specific workspaceRoot
- Prefer agent config over global for multiple agents
All tests pass.
Add 7 tests for resolveAgentConfig():
- Return undefined when no agents config exists
- Return undefined when agent id does not exist
- Return basic agent config (name, workspace, agentDir, model)
- Return agent-specific sandbox config
- Return agent-specific tools config
- Return both sandbox and tools config
- Normalize agent id
All tests pass.
Changes to defaultSandboxConfig():
- Add optional agentId parameter
- Load routing.agents[agentId].sandbox if available
- Prefer agent-specific settings over global agent.sandbox
Update callers in resolveSandboxContext() and
ensureSandboxWorkspaceForSession() to extract agentId
from sessionKey and pass it to defaultSandboxConfig().
This enables per-agent sandbox modes (e.g., main: off, family: all).
Return newly added fields from routing.agents config:
- sandbox: agent-specific sandbox configuration
- tools: agent-specific tool restrictions
This makes per-agent sandbox and tool settings accessible
to other parts of the codebase.
The /activation command now properly controls group activation mode:
- Loads session state before filtering messages
- Checks groupActivation field (from /activation command)
- Falls back to config telegram.groups requireMention setting
Previously, the bot only checked config and ignored session state,
making the /activation command appear to work but have no effect.
Changes:
- Add resolveGroupActivation() to check session before config
- Import loadSessionStore to read session state early
- Pass messageThreadId to support forum topics correctly
- Add detailed explanation of group activation modes (requireMention)
- Document /activation command (mention vs always)
- Clarify two-level access control: group allowlist + sender policy
- Add troubleshooting section for common issues
- Explain that telegram.groups creates an allowlist
- Add instructions for getting group chat ID
Fixes confusion around group setup where /activation command
updates session state but doesn't persist or take effect.
Add docs/multi-agent-sandbox-tools.md covering:
- Configuration examples (personal + restricted, work agents)
- Different sandbox modes per agent
- Tool restriction patterns (read-only, safe execution, communication-only)
- Configuration precedence rules
- Migration guide from single-agent setups
- Troubleshooting tips
Add PR_SUMMARY.md for upstream submission with:
- Feature overview and use cases
- Implementation details (49 LoC across 5 files)
- Test coverage (18 new tests, all existing tests pass)
- Backward compatibility confirmation
- Migration examples
---
Kudos to Eula, the beautiful and selfless family owl 🦉
This feature was developed to enable safe, restricted access
for family group chats while maintaining full access for
the personal assistant. Schuhu!
Add new section explaining:
- How to configure per-agent sandbox settings
- How to configure per-agent tool restrictions
- Benefits (security isolation, resource control, flexible policies)
- Link to detailed guide
Include example config showing personal assistant (no sandbox)
vs family bot (sandboxed with read-only tools).
Add 5 tests for agent-specific tool restrictions:
- Apply global tool policy when no agent-specific policy exists
- Apply agent-specific tool policy
- Allow different tool policies for different agents
- Combine global and agent-specific deny lists
- Work with sandbox tools filtering
All tests pass.
Add 6 tests for agent-specific sandbox configuration:
- Use global sandbox config when no agent-specific config exists
- Override with agent-specific sandbox mode 'off'
- Use agent-specific sandbox mode 'all'
- Use agent-specific scope
- Use agent-specific workspaceRoot
- Prefer agent config over global for multiple agents
All tests pass.
Add 7 tests for resolveAgentConfig():
- Return undefined when no agents config exists
- Return undefined when agent id does not exist
- Return basic agent config (name, workspace, agentDir, model)
- Return agent-specific sandbox config
- Return agent-specific tools config
- Return both sandbox and tools config
- Normalize agent id
All tests pass.
Changes to defaultSandboxConfig():
- Add optional agentId parameter
- Load routing.agents[agentId].sandbox if available
- Prefer agent-specific settings over global agent.sandbox
Update callers in resolveSandboxContext() and
ensureSandboxWorkspaceForSession() to extract agentId
from sessionKey and pass it to defaultSandboxConfig().
This enables per-agent sandbox modes (e.g., main: off, family: all).
Return newly added fields from routing.agents config:
- sandbox: agent-specific sandbox configuration
- tools: agent-specific tool restrictions
This makes per-agent sandbox and tool settings accessible
to other parts of the codebase.
Add source profiles anthropic:claude-cli and openai-codex:codex-cli; surface them in onboarding/configure.
Co-authored-by: pepicrft <pepicrft@users.noreply.github.com>
Use Anthropic OAuth profile email as the profile identifier when available. This fixes cases where Anthropic returns an email-based profile id rather than an explicit id field.
When queued messages come from different providers (Slack + Telegram),
process them individually instead of collecting into a single prompt.
This ensures each reply routes back to its originating provider.
- Add hasCrossProviderItems() to detect multi-provider queues
- Skip collect mode when cross-provider detected
- Preserve originatingChannel/originatingTo when collecting same-provider
When OriginatingChannel matches Surface (same provider), use normal
dispatcher. Only route via routeReply() when they differ, ensuring
cross-provider messages (e.g., Telegram queued while Slack active)
get routed back to their origin.
Implement reply routing based on OriginatingChannel/OriginatingTo fields.
This ensures replies go back to the provider where the message originated
instead of using the session's lastChannel.
Changes:
- Add OriginatingChannel/OriginatingTo fields to MsgContext (templating.ts)
- Add originatingChannel/originatingTo fields to FollowupRun (queue.ts)
- Create route-reply.ts with provider-agnostic router
- Update all providers (Telegram, Slack, Discord, Signal, iMessage)
to pass originating channel info
- Update reply.ts to pass originating channel to followupRun
- Update followup-runner.ts to use route-reply for originating channels
This addresses the issue where messages from one provider (e.g., Slack)
would receive replies on a different provider (e.g., Telegram) because
the queue used the last active dispatcher instead of the originating one.
Wrap waitForCompactionRetry() in try/catch to capture AbortError
that was escaping and bypassing the model fallback mechanism.
When a timeout fires, session.abort() causes both session.prompt()
and waitForCompactionRetry() to throw AbortError. Previously only
the prompt error was captured, allowing the compaction error to
escape to model-fallback.ts where it was immediately re-thrown
(line 199: isAbortError check), bypassing fallback model attempts.
Fixes#313
This commit fixes several issues with multi-account OAuth rotation that
were causing slow responses and inefficient account cycling.
## Changes
### 1. Fix usageStats race condition (auth-profiles.ts)
The `markAuthProfileUsed`, `markAuthProfileCooldown`, `markAuthProfileGood`,
and `clearAuthProfileCooldown` functions were using a stale in-memory store
passed as a parameter. Long-running sessions would overwrite usageStats
updates from concurrent sessions when saving.
**Fix:** Re-read the store from disk before each update to get fresh
usageStats from other sessions, then merge the update.
### 2. Capture AbortError from waitForCompactionRetry (pi-embedded-runner.ts)
When a request timed out, `session.abort()` was called which throws an
`AbortError`. The code structure was:
```javascript
try {
await session.prompt(params.prompt);
} catch (err) {
promptError = err; // Catches AbortError here
}
await waitForCompactionRetry(); // But THIS also throws AbortError!
```
The second `AbortError` from `waitForCompactionRetry()` escaped and
bypassed the rotation/fallback logic entirely.
**Fix:** Wrap `waitForCompactionRetry()` in its own try/catch to capture
the error as `promptError`, enabling proper timeout handling.
Root cause analysis and fix proposed by @erikpr1994 in #313.
Fixes#313
### 3. Fail fast on 429 rate limits (pi-ai patch)
The pi-ai library was retrying 429 errors up to 3 times with exponential
backoff before throwing. This meant a rate-limited account would waste
30+ seconds retrying before our rotation code could try the next account.
**Fix:** Patch google-gemini-cli.js to:
- Throw immediately on first 429 (no retries)
- Not catch and retry 429 errors in the network error handler
This allows the caller to rotate to the next account instantly on rate limit.
Note: We submitted this fix upstream (https://github.com/badlogic/pi-mono/pull/504)
but it was closed without merging. Keeping as a local patch for now.
## Testing
With 6 Antigravity accounts configured:
- Accounts rotate properly based on lastUsed (round-robin)
- 429s trigger immediate rotation to next account
- usageStats persist correctly across concurrent sessions
- Cooldown tracking works as expected
## Before/After
**Before:** Multiple 429 retries on same account, 30-90s delays
**After:** Instant rotation on 429, responses in seconds
Discord voice messages have empty `content` with the audio in attachments.
The nullish coalescing operator (`??`) doesn't fall through on empty strings,
so voice messages were being dropped as 'empty content'.
Changed to logical OR (`||`) so empty string falls through to media placeholder.
Adds a /stop command that:
- Can interrupt a running agent session mid-execution
- Works in both DMs and group chats (including forum topics)
- Uses grammy's bot.command() to run before the main message handler
- Returns status: stopped, stop requested, or nothing running
Also fixes session key lookup in pi-embedded-runner to use sessionKey
instead of sessionId, ensuring /stop finds the correct active run.
Add skill for Himalaya (https://github.com/pimalaya/himalaya), a CLI
email client supporting IMAP, SMTP, Notmuch, and Sendmail backends.
Includes:
- SKILL.md with common operations (list, read, reply, forward, send)
- Configuration reference for Gmail, iCloud, and generic IMAP/SMTP
- MML (MIME Meta Language) composition guide for attachments
Tested with iCloud IMAP account - verified folder listing, email
reading, and sending work correctly.
- Added `isError` property to `EmbeddedPiRunResult` and reply items to indicate error states.
- Updated error handling in `runReplyAgent` to provide more informative messages for specific socket connection errors.
- tabs.ts now uses getProfileContext like other routes
- browser-tool threads profile param through all actions
- add tests for profile query param on /tabs endpoints
- update docs with browser tool profile parameter
- Onboarding now writes auth profiles under ~/.clawdbot/agents/main/agent so the gateway sees credentials on first start.
- Hardened onboarding test to ignore legacy env vars.
Thanks @minghinmatthewlam!
Bun's WebSocket implementation doesn't fully support Playwright's CDP
connection because Playwright bundles its own 'ws' module. This causes
connectOverCDP to timeout.
The patch makes Playwright use the native 'ws' module when running
under Bun, which works with Bun's WebSocket shim.
Fixes browser snapshot/act timeouts after PR #278 (tsx → bun migration).
Ref: https://github.com/oven-sh/bun/issues/9911
- tabs.ts now uses getProfileContext like other routes
- browser-tool threads profile param through all actions
- add tests for profile query param on /tabs endpoints
- update docs with browser tool profile parameter
* fix(slack): use named import for @slack/bolt App class
The default import `import bolt from '@slack/bolt'` followed by
`const { App } = bolt` doesn't work correctly in Bun due to ESM/CJS
interop issues. The default export comes through as a function rather
than the module object.
Switching to a named import `import { App } from '@slack/bolt'`
resolves the issue and allows the Slack provider to start successfully.
* fix(slack): align Bolt mock with named App export
---------
Co-authored-by: Peter Steinberger <steipete@gmail.com>
Detect the Gemini API error 'function call turn comes immediately after
a user turn or after a function response turn' which indicates corrupted
session history.
When detected:
- Delete the corrupted transcript file
- Remove the session entry from the store
- Return a user-friendly message asking them to retry
This prevents the error loop where every subsequent message fails with
the same error until manual intervention.
Fixes#296
* fix(auth): prioritize round-robin over lastGood for multi-account rotation
When multiple OAuth accounts are configured, the round-robin rotation was
not working because lastGood was always prioritized, defeating the sort by
lastUsed.
Changes:
- Remove lastGood prioritization in resolveAuthProfileOrder
- Always apply orderProfilesByMode (sorts by lastUsed, oldest first)
- Only respect configuredOrder when explicitly set in config
- preferredProfile still takes priority for explicit user choice
Tested with 2 Google Antigravity accounts - verified alternating usage.
Follow-up to PR #269.
* style: fix formatting
The chat view was starting at the top showing oldest messages instead of
scrolling to the bottom to show the latest messages (like WhatsApp).
Root causes:
1. scheduleChatScroll() was called without force flag in refreshActiveTab()
2. The scroll was targeting .chat-thread element which has overflow:visible
and doesn't actually scroll - the window scrolls instead
Fixes:
- Pass force flag (!chatHasAutoScrolled) when loading chat tab
- Wait for Lit updateComplete before scrolling to ensure DOM is ready
- Scroll the window instead of the .chat-thread container
- Use behavior: 'instant' for immediate scroll without animation
- Add TelegramLocation, TelegramVenue, and TelegramMessageWithLocation types
- Add formatLocationMessage() to convert location/venue shares to text
- Add extractLocationData() for structured location access in ctxPayload
- Handle both raw location pins and venue shares (places with names)
- Include location in reply-to context for quoted messages
Location messages now appear as:
- [Location: lat, lon ±accuracy] for raw pins
- [Venue: Name - Address (lat, lon)] for places
ctxPayload includes LocationLat, LocationLon, LocationAccuracy,
VenueName, and VenueAddress fields for programmatic access.
New section explaining how to periodically review daily memory files
and update MEMORY.md with distilled learnings. Like a human reviewing
their journal and updating their mental model.
Antigravity rate limits cause requests to hang indefinitely rather than
returning 429 errors. This change detects timeouts and treats them as
potential rate limits:
- Added timedOut flag to track timeout-triggered aborts
- Timeout now triggers profile cooldown + rotation
- Logs: "Profile X timed out (possible rate limit). Trying next account..."
This ensures automatic failover when Antigravity hangs due to rate limiting.
- Updated session start to include MEMORY.md loading for main sessions
- Added 🧠 MEMORY.md section explaining:
- Only load in main sessions (direct with human), not shared contexts
- Security boundary: personal context shouldn't leak to strangers
- Can freely read/edit/update in main sessions
- Write significant events, thoughts, decisions, opinions
- Curated memory vs raw daily logs
This gives new agents proper long-term memory that's secure and personal.
Adds usage tracking to auth profiles for automatic rotation:
- ProfileUsageStats type with lastUsed, cooldownUntil, errorCount
- markAuthProfileUsed(): tracks successful usage, resets errors
- markAuthProfileCooldown(): applies exponential backoff (1/5/25/60min)
- isProfileInCooldown(): checks if profile should be skipped
- orderProfilesByMode(): now sorts by lastUsed (oldest first)
On auth/rate-limit failures, profiles are marked for cooldown before
rotation. On success, usage is recorded for round-robin ordering.
This enables automatic load distribution across multiple accounts
(e.g., Antigravity 5-hour rate limit windows).
Changes writeOAuthCredentials and applyAuthProfileConfig calls to use
the email from OAuth response as part of the profile ID instead of
hardcoded ":default".
This enables multiple accounts per provider - each login creates a
separate profile (e.g., google-antigravity:user@gmail.com) instead
of overwriting the same :default profile.
Affected files:
- src/commands/onboard-auth.ts (generic writeOAuthCredentials)
- src/commands/configure.ts (Antigravity flow)
- src/wizard/onboarding.ts (Antigravity flow)
* fix(ui): add anyOf/oneOf support in config form
- Handle literal unions as dropdowns with type preservation
- Handle primitive unions (string|number, boolean|string) as text inputs
- Unwrap single-variant optional types
- Fix enum handler to preserve types via index-based values
- Update normalizeUnion to support primitive unions in schema analysis
- Exclude allOf from union normalization (stays unsupported)
Fields like Thinking Default, Allow From, Memory now render properly
instead of showing 'unsupported schema node' errors.
* UI: fix enum placeholder collision
* Docs: update changelog for PR #268
---------
Co-authored-by: Shadow <hi@shadowing.dev>
* fix: Gemini stops working after one message in a session
* fix: small issue in test file
* test: cover google role-merge behavior
---------
Co-authored-by: Peter Steinberger <steipete@gmail.com>
When replying to a message in a Slack thread, the response now stays
in the thread instead of going to the channel root.
Only threads replies when the incoming message was already in a thread
(has thread_ts). Top-level messages get top-level replies.
Fixes#250
what: default bash PATH to process.env.PATH
why: ensure Nix-provided tools on PATH inside sessions
tests: not run
Co-authored-by: Peter Steinberger <steipete@gmail.com>
The manifest was missing scopes required for conversations.open API,
which is used to get DM channel IDs for replies.
Added scopes:
- im:write (required for DM replies)
- im:read (list DM conversations)
- mpim:write (reply to multi-person DMs)
- mpim:read (list MPDMs)
- groups:write (private channel interactions)
- groups:read (list private channels)
Without im:write, the example config (dm.enabled: true) cannot
actually reply to DMs - fails with missing_scope error.
Co-authored-by: Manuel Hettich <17690367+ManuelHettich@users.noreply.github.com>
Co-authored-by: Peter Steinberger <steipete@gmail.com>
* cron: skip delivery for HEARTBEAT_OK responses
When an isolated cron job has deliver:true, skip message delivery if the
response is just HEARTBEAT_OK (or contains HEARTBEAT_OK at edges with
short remaining content <= 30 chars). This allows cron jobs to silently
ack when nothing to report but still deliver actual content when there
is something meaningful to say.
Media is still delivered even if text is HEARTBEAT_OK, since the
presence of media indicates there's something to share.
* fix(heartbeat): make ack padding configurable
* chore(deps): update to latest
---------
Co-authored-by: Josh Lehman <josh@martian.engineering>
When an isolated cron job has deliver:true, skip message delivery if the
response is just HEARTBEAT_OK (or contains HEARTBEAT_OK at edges with
short remaining content <= 30 chars). This allows cron jobs to silently
ack when nothing to report but still deliver actual content when there
is something meaningful to say.
Media is still delivered even if text is HEARTBEAT_OK, since the
presence of media indicates there's something to share.
The SDK's tools parameter only accepts built-in tools (read, bash, edit, write).
Custom clawdbot tools (browser, canvas, nodes, cron, etc.) were being filtered
out, causing 'Tool not found' errors at runtime.
Split tools into built-in and custom, passing them via the correct parameters.
The outer fence (4 backticks) was closing prematurely after the bash
example, leaving the rest of the template (Feature intent through
Submitted by Razor) rendered as prose instead of inside the code block.
Fixed by moving the closing fence to the end of the full template.
**Important:** Always wait for the Telegram verification message before proceeding. If the agent doesn't respond, troubleshoot the gateway or model configuration before pushing.
Upstream updates may introduce Swift 6.2 / macOS 26 SDK incompatibilities. Use analyze-mode for systematic debugging:
### Analyze-Mode Investigation
```bash
# Gather context with parallel agents
morph-mcp_warpgrep_codebase_search search_string="Find deprecated FileManager.default and Thread.isMainThread usages in Swift files"repo_path="/Volumes/Main SSD/Developer/clawdis"
morph-mcp_warpgrep_codebase_search search_string="Locate Peekaboo submodule and macOS app Swift files with concurrency issues"repo_path="/Volumes/Main SSD/Developer/clawdis"
Common issue: `fetch.preconnect` type mismatch. Fix by using `FetchLike` type instead of `typeof fetch`.
### macOS App Crashes on Launch
Usually resource bundle mismatch. Full rebuild required:
```bash
cd apps/macos && rm -rf .build .swiftpm
./scripts/restart-mac.sh
```
### Patch Failures
```bash
# Check patch status
pnpm install 2>&1| grep -i patch
# If patches fail, they may need updating for new dep versions
# Check patches/ directory against package.json patchedDependencies
```
### Swift 6.2 / macOS 26 SDK Build Failures
**Symptoms:** Build fails with deprecation warnings about `FileManager.default` or `Thread.isMainThread`
**Search-Mode Investigation:**
```bash
# Exhaustive search for deprecated APIs
morph-mcp_warpgrep_codebase_search search_string="Find all Swift files using deprecated FileManager.default or Thread.isMainThread"repo_path="/Volumes/Main SSD/Developer/clawdis"
- GitHub issues/comments/PR comments: use literal multiline strings or `-F - <<'EOF'` (or $'...') for real newlines; never embed "\\n".
## Project Structure & Module Organization
- Source code: `src/` (CLI wiring in `src/cli`, commands in `src/commands`, web provider in `src/provider-web.ts`, infra in `src/infra`, media pipeline in `src/media`).
- Tests: colocated `*.test.ts`.
- Docs: `docs/` (images, queue, Pi config). Built output lives in `dist/`.
- Plugins/extensions: live under `extensions/*` (workspace packages). Keep plugin-only deps in the extension `package.json`; do not add them to the root `package.json` unless core uses them.
- Plugins: install runs `npm install --omit=dev` in plugin dir; runtime deps must live in `dependencies`. Avoid `workspace:*` in `dependencies` (npm install breaks); put `clawdbot` in `devDependencies` or `peerDependencies` instead (runtime resolves `clawdbot/plugin-sdk` via jiti alias).
- Installers served from `https://clawd.bot/*`: live in the sibling repo `../clawd.bot` (`public/install.sh`, `public/install-cli.sh`, `public/install.ps1`).
- dev: moving head on `main` (no tag; git checkout main).
## Testing Guidelines
- Framework: Vitest with V8 coverage thresholds (70% lines/branches/functions/statements).
- Naming: match source names with `*.test.ts`; e2e in `*.e2e.test.ts`.
- Run `pnpm test` (or `pnpm test:coverage`) before pushing when you touch logic.
- Do not set test workers above 16; tried already.
- Live tests (real keys): `CLAWDBOT_LIVE_TEST=1 pnpm test:live` (Clawdbot-only) or `LIVE=1 pnpm test:live` (includes provider live tests). Docker: `pnpm test:docker:live-models`, `pnpm test:docker:live-gateway`. Onboarding Docker E2E: `pnpm test:docker:onboard`.
- Full kit + what’s covered: `docs/testing.md`.
- Pure test additions/fixes generally do **not** need a changelog entry unless they alter user-facing behavior or the user asks for one.
- Mobile: before using a simulator, check for connected real devices (iOS + Android) and prefer them when available.
@@ -29,50 +72,88 @@
- Create commits with `scripts/committer "<msg>" <file...>`; avoid manual `git add`/`git commit` so staging stays scoped.
- Follow concise, action-oriented commit messages (e.g., `CLI: add verbose flag to send`).
- Group related changes; avoid bundling unrelated refactors.
- Changelog workflow: keep latest released version at top (no `Unreleased`); after publishing, bump version and start a new top section.
- PRs should summarize scope, note testing performed, and mention any user-facing changes or new flags.
- PR review flow: when given a PR link, review via `gh pr view`/`gh pr diff` and do **not** change branches.
- PR review calls: prefer a single `gh pr view --json ...` to batch metadata/comments; run `gh pr diff` only when needed.
- Before starting a review when a GH Issue/PR is pasted: run `git pull`; if there are local changes or unpushed commits, stop and alert the user before reviewing.
- Goal: merge PRs. Prefer **rebase** when commits are clean; **squash** when history is messy.
- PR merge flow: create a temp branch from `main`, merge the PR branch into it (prefer squash unless commit history is important; use rebase/merge when it is). Always try to merge the PR unless it’s truly difficult, then use another approach. If we squash, add the PR author as a co-contributor. Apply fixes, add changelog entry (include PR # + thanks), run full gate before the final commit, commit, merge back to `main`, delete the temp branch, and end on `main`.
- If you review a PR and later do work on it, land via merge/squash (no direct-main commits) and always add the PR author as a co-contributor.
- When working on a PR: add a changelog entry with the PR number and thank the contributor.
- When working on an issue: reference the issue in the changelog entry.
- When merging a PR: leave a PR comment that explains exactly what we did and include the SHA hashes.
- When merging a PR from a new contributor: add their avatar to the README “Thanks to all clawtributors” thumbnail list.
- After merging a PR: run `bun scripts/update-clawtributors.ts` if the contributor is missing, then commit the regenerated README.
## Shorthand Commands
-`sync`: if working tree is dirty, commit all changes (pick a sensible Conventional Commit message), then `git pull --rebase`; if rebase conflicts and cannot resolve, stop; otherwise `git push`.
- **Landing mode:** create an integration branch from `main`, bring in PR commits (**prefer rebase** for linear history; **merge allowed** when complexity/conflicts make it safer), apply fixes, add changelog (+ thanks + PR #), run full gate **locally before committing** (`pnpm lint && pnpm build && pnpm test`), commit, merge back to `main`, then `git switch main` (never stay on a topic branch after landing). Important: contributor needs to be in git graph after this!
## Security & Configuration Tips
- Web provider stores creds at `~/.clawdbot/credentials/`; rerun `clawdbot login` if logged out.
- Pi sessions live under `~/.clawdbot/sessions/` by default; the base directory is not configurable.
- Environment variables: see `~/.profile`.
- Never commit or publish real phone numbers, videos, or live configuration values. Use obviously fake placeholders in docs, tests, and examples.
- Release flow: always read `docs/reference/RELEASING.md` and `docs/platforms/mac/release.md` before any release work; do not ask routine questions once those docs answer them.
## Troubleshooting
- Rebrand/migration issues or legacy config/service warnings: run `clawdbot doctor` (see `docs/gateway/doctor.md`).
## Agent-Specific Notes
-Gateway currently runs only as the menubar app (launchctl shows `application.com.steipete.clawdbot.debug.*`), there is no separate LaunchAgent/helper label installed. Restart via the Clawdbot Mac app or `scripts/restart-mac.sh`; to verify/kill use `launchctl print gui/$UID | grep clawdbot` rather than expecting `com.steipete.clawdbot`. **When debugging on macOS, start/stop the gateway via the app, not ad-hoc tmux sessions; kill any temporary tunnels before handoff.**
-macOS logs: use `./scripts/clawlog.sh` (aka `vtlog`) to query unified logs for subsystem `com.steipete.clawdbot`; it supports follow/tail/category filters and expects passwordless sudo for `/usr/bin/log`.
-Also read the shared guardrails at `~/Projects/oracle/AGENTS.md` and `~/Projects/agent-scripts/AGENTS.MD` before making changes; align with any cross-repo rules noted there.
-Vocabulary: "makeup" = "mac app".
-Never edit `node_modules` (global/Homebrew/npm/git installs too). Updates overwrite. Skill notes go in `tools.md` or `AGENTS.md`.
-When working on a GitHub Issue or PR, print the full URL at the end of the task.
- When answering questions, respond with high-confidence answers only: verify in code; do not guess.
- Never update the Carbon dependency.
- Any dependency with `pnpm.patchedDependencies` must use an exact version (no `^`/`~`).
- Patching dependencies (pnpm patches, overrides, or vendored changes) requires explicit approval; do not do this by default.
- Gateway currently runs only as the menubar app; there is no separate LaunchAgent/helper label installed. Restart via the Clawdbot Mac app or `scripts/restart-mac.sh`; to verify/kill use `launchctl print gui/$UID | grep clawdbot` rather than assuming a fixed label. **When debugging on macOS, start/stop the gateway via the app, not ad-hoc tmux sessions; kill any temporary tunnels before handoff.**
- macOS logs: use `./scripts/clawlog.sh` to query unified logs for the Clawdbot subsystem; it supports follow/tail/category filters and expects passwordless sudo for `/usr/bin/log`.
- If shared guardrails are available locally, review them; otherwise follow this repo's guidance.
- SwiftUI state management (iOS/macOS): prefer the `Observation` framework (`@Observable`, `@Bindable`) over `ObservableObject`/`@StateObject`; don’t introduce new `ObservableObject` unless required for compatibility, and migrate existing usages when touching related code.
- Connection providers: when adding a new connection, update every UI surface and docs (macOS app, web UI, mobile if applicable, onboarding/overview docs) and add matching status + configuration forms so provider lists and settings stay in sync.
- **Restart apps:** “restart iOS/Android apps” means rebuild (recompile/install) and relaunch, not just kill/launch.
- **Device checks:** before testing, verify connected real devices (iOS/Android) before reaching for simulators/emulators.
- iOS Team ID lookup: `security find-identity -p codesigning -v` → use Apple Development (…) TEAMID. Fallback: `defaults read com.apple.dt.Xcode IDEProvisioningTeamIdentifiers`.
- A2UI bundle hash: `src/canvas-host/a2ui/.bundle.hash` is auto-generated; regenerate via `pnpm canvas:a2ui:bundle` (or `scripts/bundle-a2ui.sh`) instead of manual conflict resolution.
-Notary key file lives at `~/Library/CloudStorage/Dropbox/Backup/AppStore/AuthKey_NJF3NFGTS3.p8` (Sparkle keys live under `~/Library/CloudStorage/Dropbox/Backup/Sparkle`).
-**Multi-agent safety:** do **not** create/apply/drop `git stash` entries unless Peter explicitly asks (this includes `git pull --rebase --autostash`). Assume other agents may be working; keep unrelated WIP untouched and avoid cross-cutting state changes.
- **Multi-agent safety:** when Peter says "push", you may `git pull --rebase` to integrate latest changes (never discard other agents' work). When Peter says "commit", scope to your changes only. When Peter says "commit all", commit everything in grouped chunks.
- **Multi-agent safety:** do **not** create/remove/modify `git worktree` checkouts (or edit `.worktrees/*`) unless Peter explicitly asks.
- **Multi-agent safety:** do **not**switch branches / checkout a different branch unless Peter explicitly asks.
- A2UI bundle hash: `src/canvas-host/a2ui/.bundle.hash` is auto-generated; ignore unexpected changes, and only regenerate via `pnpm canvas:a2ui:bundle` (or `scripts/bundle-a2ui.sh`) when needed. Commit the hash as a separate commit.
-Release signing/notary keys are managed outside the repo; follow internal release docs.
-Notary auth env vars (`APP_STORE_CONNECT_ISSUER_ID`, `APP_STORE_CONNECT_KEY_ID`, `APP_STORE_CONNECT_API_KEY_P8`) are expected in your environment (per internal release docs).
- **Multi-agent safety:** do **not** create/apply/drop `git stash` entries unless explicitly requested (this includes `git pull --rebase --autostash`). Assume other agents may be working; keep unrelated WIP untouched and avoid cross-cutting state changes.
- **Multi-agent safety:** when the user says "push", you may `git pull --rebase` to integrate latest changes (never discard other agents' work). When the user says "commit", scope to your changes only. When the user says "commit all", commit everything in grouped chunks.
- **Multi-agent safety:** do **not** switch branches / check out a different branch unless explicitly requested.
- **Multi-agent safety:** running multiple agents is OK as long as each agent has its own session.
-When asked to open a “session” file, open the Pi session logs under `~/.clawdbot/sessions/*.jsonl` (newest unless a specific ID is given), not the default `sessions.json`. If logs are needed from Mac Studio, SSH via Tailscale and read the same path there.
-Menubar dimming + restart flow mirrors Trimmy: use `scripts/restart-mac.sh` (kills all Clawdbot variants, runs `swift build`, packages, relaunches). Icon dimming depends on MenuBarExtraAccess wiring in AppMain; keep `appearsDisabled` updates intact when touching the status item.
-**Multi-agent safety:** when you see unrecognized files, keep going; focus on your changes and commit only those.
-Lint/format churn:
- If staged+unstaged diffs are formatting-only, auto-resolve without asking.
- If commit/push already requested, auto-stage and include formatting-only follow-ups in the same commit (or a tiny follow-up commit if needed), no extra confirmation.
- Only ask when changes are semantic (logic/data/behavior).
- Lobster seam: use the shared CLI palette in `src/terminal/palette.ts` (no hardcoded colors); apply palette to onboarding/config prompts and other TTY UI output as needed.
- **Multi-agent safety:** focus reports on your edits; avoid guard-rail disclaimers unless truly blocked; when multiple agents touch the same file, continue if safe; end with a brief “other files present” note only if relevant.
- Bug investigations: read source code of relevant npm dependencies and all related local code before concluding; aim for high-confidence root cause.
- Code style: add brief comments for tricky logic; keep files under ~500 LOC when feasible (split/refactor as needed).
- Tool schema guardrails (google-antigravity): avoid `Type.Union` in tool input schemas; no `anyOf`/`oneOf`/`allOf`. Use `stringEnum`/`optionalStringEnum` (Type.Unsafe enum) for string lists, and `Type.Optional(...)` instead of `... | null`. Keep top-level tool schema as `type: "object"` with `properties`.
- Tool schema guardrails: avoid raw `format` property names in tool schemas; some validators treat `format` as a reserved keyword and reject the schema.
- When asked to open a “session” file, open the Pi session logs under `~/.clawdbot/agents/<agentId>/sessions/*.jsonl` (use the `agent=<id>` value in the Runtime line of the system prompt; newest unless a specific ID is given), not the default `sessions.json`. If logs are needed from another machine, SSH via Tailscale and read the same path there.
- Do not rebuild the macOS app over SSH; rebuilds must be run directly on the Mac.
- Never send streaming/partial replies to external messaging surfaces (WhatsApp, Telegram); only final replies should be delivered there. Streaming/tool events may still go to internal UIs/control channel.
- Voice wake forwarding tips:
- Command template should stay `clawdbot-mac agent --message "${text}" --thinking low`; `VoiceWakeForwarder` already shell-escapes `${text}`. Don’t add extra quotes.
- launchd PATH is minimal; ensure the app’s launch agent sets PATH to include`/opt/homebrew/bin:/usr/local/bin:/usr/bin:/bin:/usr/sbin:/sbin:/Users/steipete/Library/pnpm` so `pnpm`/`clawdbot` binaries resolve when invoked via `clawdbot-mac`.
- For manual `clawdbot send` messages that include `!`, use the heredoc pattern noted below to avoid the Bash tool’s escaping.
- launchd PATH is minimal; ensure the app’s launch agent PATH includes standard system paths plus your pnpm bin (typically `$HOME/Library/pnpm`) so `pnpm`/`clawdbot` binaries resolve when invoked via `clawdbot-mac`.
- For manual `clawdbot message send` messages that include `!`, use the heredoc pattern noted below to avoid the Bash tool’s escaping.
- Release guardrails: do not change version numbers without operator’s explicit consent; always ask permission before running any npm publish/release step.
## Exclamation Mark Escaping Workaround
The Claude Code Bash tool escapes `!` to `\\!` in command arguments. When using `clawdbot send` with messages containing exclamation marks, use heredoc syntax:
**Clawdbot** is a *personal AI assistant* you run on your own devices.
It answers you on the surfaces you already use (WhatsApp, Telegram, Discord, iMessage, WebChat), can speak and listen on macOS/iOS, and can render a live Canvas you control. The Gateway is just the control plane — the product is the assistant.
It answers you on the channels you already use (WhatsApp, Telegram, Slack, Discord, Signal, iMessage, Microsoft Teams, WebChat), plus extension channels like BlueBubbles, Matrix, Zalo, and Zalo Personal. It can speak and listen on macOS/iOS/Android, and can render a live Canvas you control. The Gateway is just the control plane — the product is the assistant.
If you want a personal, single-user assistant that feels local, fast, and always-on, this is it.
Preferred setup: run the onboarding wizard (`clawdbot onboard`). It walks through gateway, workspace, providers, and skills. The CLI wizard is the recommended path and works on **macOS, Windows, and Linux**.
Preferred setup: run the onboarding wizard (`clawdbot onboard`). It walks through gateway, workspace, channels, and skills. The CLI wizard is the recommended path and works on **macOS, Linux, and Windows (via WSL2; strongly recommended)**.
Works with npm, pnpm, or bun.
New install? Start here: [Getting started](https://docs.clawd.bot/start/getting-started)
Using Claude Pro/Max subscription? See `docs/onboarding.md` for the Anthropic OAuth setup.
Model note: while any model is supported, I strongly recommend **Anthropic Pro/Max (100/200) + Opus 4.5** for long‑context strength and better prompt‑injection resistance. See [Onboarding](https://docs.clawd.bot/start/onboarding).
- Auth profile rotation (OAuth vs API keys) + fallbacks: [Model failover](https://docs.clawd.bot/concepts/model-failover)
## Install (recommended)
Runtime: **Node ≥22**.
```bash
npm install -g clawdbot@latest
# or: pnpm add -g clawdbot@latest
clawdbot onboard --install-daemon
```
The wizard installs the Gateway daemon (launchd/systemd user service) so it stays running.
## Quick start (TL;DR)
Runtime: **Node ≥22**.
Full beginner guide (auth, pairing, channels): [Getting started](https://docs.clawd.bot/start/getting-started)
```bash
clawdbot onboard --install-daemon
clawdbot gateway --port 18789 --verbose
# Send a message
clawdbot message send --to +1234567890 --message "Hello from Clawdbot"
# Talk to the assistant (optionally deliver back to any connected channel: WhatsApp/Telegram/Slack/Discord/Signal/iMessage/BlueBubbles/Microsoft Teams/Matrix/Zalo/Zalo Personal/WebChat)
clawdbot agent --message "Ship checklist" --thinking high
```
Upgrading? [Updating guide](https://docs.clawd.bot/install/updating) (and run `clawdbot doctor`).
## Development channels
- **stable**: tagged releases (`vYYYY.M.D` or `vYYYY.M.D-<patch>`), npm dist-tag `latest`.
- **beta**: prerelease tags (`vYYYY.M.D-beta.N`), npm dist-tag `beta` (macOS app may be missing).
- **dev**: moving head of `main`, npm dist-tag `dev` (when published).
pnpm ui:build # auto-installs UI deps on first run
pnpm build
pnpm clawdbot onboard --install-daemon
# Dev loop (auto-reload on TS changes)
pnpm gateway:watch
```
Note: `pnpm clawdbot ...` runs TypeScript directly (via `tsx`). `pnpm build` produces `dist/` for running via Node / the packaged `clawdbot` binary.
## Security defaults (DM access)
Clawdbot connects to real messaging surfaces. Treat inbound DMs as **untrusted input**.
Full security guide: [Security](https://docs.clawd.bot/gateway/security)
Default behavior on Telegram/WhatsApp/Signal/iMessage/Microsoft Teams/Discord/Slack:
- **DM pairing** (`dmPolicy="pairing"` / `channels.discord.dm.policy="pairing"` / `channels.slack.dm.policy="pairing"`): unknown senders receive a short pairing code and the bot does not process their message.
- Approve with: `clawdbot pairing approve <channel> <code>` (then the sender is added to a local allowlist store).
- Public inbound DMs require an explicit opt-in: set `dmPolicy="open"` and include `"*"` in the channel allowlist (`allowFrom` / `channels.discord.dm.allowFrom` / `channels.slack.dm.allowFrom`).
Run `clawdbot doctor` to surface risky/misconfigured DM policies.
## Highlights
- **Local-first Gateway** — single control plane for sessions, providers, tools, and events.
- **[Companion apps](https://docs.clawd.bot/platforms/macos)** — macOS menu bar app + iOS/Android [nodes](https://docs.clawd.bot/nodes).
- **[Onboarding](https://docs.clawd.bot/start/wizard) + [skills](https://docs.clawd.bot/tools/skills)** — wizard-driven setup with bundled/managed/workspace skills.
## Star History
[](https://www.star-history.com/#clawdbot/clawdbot&type=date&legend=top-left)
## Everything we built so far
### Core platform
- Gateway WS control plane with sessions, presence, config, cron, webhooks, control UI, and Canvas host.
- CLI surface: gateway, agent, send, wizard, doctor/update, and TUI.
- Pi agent runtime in RPC mode with tool streaming and block streaming.
- Session model: `main` for direct chats, group isolation, activation modes, queue modes, reply-back.
-[Gateway WS control plane](https://docs.clawd.bot/gateway) with sessions, presence, config, cron, webhooks, [Control UI](https://docs.clawd.bot/web), and [Canvas host](https://docs.clawd.bot/platforms/mac/canvas#canvas-a2ui).
-[CLI surface](https://docs.clawd.bot/tools/agent-send): gateway, agent, send, [wizard](https://docs.clawd.bot/start/wizard), and [doctor](https://docs.clawd.bot/gateway/doctor).
-[Pi agent runtime](https://docs.clawd.bot/concepts/agent) in RPC mode with tool streaming and block streaming.
-[Session model](https://docs.clawd.bot/concepts/session): `main` for direct chats, group isolation, activation modes, queue modes, reply-back. Group rules: [Groups](https://docs.clawd.bot/concepts/groups).
- Tool names drop the `clawdbot_` prefix (`browser`, `canvas`, `nodes`, `cron`, `gateway`).
- Bash tool removed `stdinMode: "pty"` support (use tmux for real TTYs).
- Primary session key is fixed to `main` (or `global` for global scope).
## Project rename + changelog format
Clawdis → Clawdbot. The rename touched every surface, path, and bundle ID. To make that transition explicit, releases now use **date-based versions** (`YYYY.M.D`), and the changelog is compressed into milestone summaries instead of long semver trains. Full detail still lives in git history and the docs.
-[Control UI](https://docs.clawd.bot/web) + [WebChat](https://docs.clawd.bot/web/webchat) served directly from the Gateway.
-[Tailscale Serve/Funnel](https://docs.clawd.bot/gateway/tailscale) or [SSH tunnels](https://docs.clawd.bot/gateway/remote) with token/password auth.
-[Nix mode](https://docs.clawd.bot/install/nix) for declarative config; [Docker](https://docs.clawd.bot/install/docker)-based installs.
- **[Gateway WebSocket network](https://docs.clawd.bot/concepts/architecture)** — single WS control plane for clients, tools, and events (plus ops: [Gateway runbook](https://docs.clawd.bot/gateway)).
- **[Tailscale exposure](https://docs.clawd.bot/gateway/tailscale)** — Serve/Funnel for the Gateway dashboard + WS (remote access: [Remote](https://docs.clawd.bot/gateway/remote)).
- **[Browser control](https://docs.clawd.bot/tools/browser)** — clawd‑managed Chrome/Chromium with CDP control.
- **[Nodes](https://docs.clawd.bot/nodes)** — Canvas, camera snap/clip, screen record, `location.get`, notifications, plus macOS‑only `system.run`/`system.notify`.
## Tailscale access (Gateway dashboard)
Clawdbot can auto-configure Tailscale **Serve** (tailnet-only) or **Funnel** (public) while the Gateway stays bound to loopback. Configure `gateway.tailscale.mode`:
-`off`: no Tailscale automation (default).
-`serve`: tailnet-only HTTPS via `tailscale serve` (uses Tailscale identity headers by default).
-`funnel`: public HTTPS via `tailscale funnel` (requires shared password auth).
Notes:
-`gateway.bind` must stay `loopback` when Serve/Funnel is enabled (Clawdbot enforces this).
- Serve can be forced to require a password by setting `gateway.auth.mode: "password"` or `gateway.auth.allowTailscale: false`.
- Funnel refuses to start unless `gateway.auth.mode: "password"` is set.
- Optional: `gateway.tailscale.resetOnExit` to undo Serve/Funnel on shutdown.
It’s perfectly fine to run the Gateway on a small Linux instance. Clients (macOS app, CLI, WebChat) can connect over **Tailscale Serve/Funnel** or **SSH tunnels**, and you can still pair device nodes (macOS/iOS/Android) to execute device‑local actions when needed.
- **Gateway host** runs the exec tool and channel connections by default.
- **Device nodes** run device‑local actions (`system.run`, camera, screen recording, notifications) via `node.invoke`.
In short: exec runs where the Gateway lives; device actions run where the device lives.
The macOS app can run in **node mode** and advertises its capabilities + permission map over the Gateway WebSocket (`node.list` / `node.describe`). Clients can then execute local actions via `node.invoke`:
-`system.run` runs a local command and returns stdout/stderr/exit code; set `needsScreenRecording: true` to require screen-recording permission (otherwise you’ll get `PERMISSION_MISSING`).
-`system.notify` posts a user notification and fails if notifications are denied.
-`canvas.*`, `camera.*`, `screen.record`, and `location.get` are also routed via `node.invoke` and follow TCC permission status.
Elevated bash (host permissions) is separate from macOS TCC:
- Use `/elevated on|off` to toggle per‑session elevated access when enabled + allowlisted.
- Gateway persists the per‑session toggle via `sessions.patch` (WS method) alongside `thinkingLevel`, `verboseLevel`, `model`, `sendPolicy`, and `groupActivation`.
-`/restart` — restart the gateway (owner-only in groups)
-`/activation mention|always` — group activation toggle (groups only)
## Architecture
## Apps (optional)
### TypeScript Gateway (src/gateway/server.ts)
- **Single HTTP+WS server** on `ws://127.0.0.1:18789` (bind policy: loopback/lan/tailnet/auto). The first frame must be `connect`; AJV validates frames against TypeBox schemas (`src/gateway/protocol`).
- **Single source of truth** for sessions, providers, cron, voice wake, and presence. Methods cover `send`, `agent`, `chat.*`, `sessions.*`, `config.*`, `cron.*`, `voicewake.*`, `node.*`, `system-*`, `wake`.
- **Events + snapshot**: handshake returns a snapshot (presence/health) and declares event types; runtime events include `agent`, `chat`, `presence`, `tick`, `health`, `heartbeat`, `cron`, `node.pair.*`, `voicewake.changed`, `shutdown`.
- **Idempotency & safety**: `send`/`agent`/`chat.send` require idempotency keys with a TTL cache (5 min, cap 1000) to avoid double‑sends on reconnects; payload sizes are capped per connection.
- **Bridge for nodes**: optional TCP bridge (`src/infra/bridge/server.ts`) is newline‑delimited JSON frames (`hello`, pairing, RPC, `invoke`); node connect/disconnect is surfaced into presence.
- **Control UI + Canvas Host**: HTTP serves Control UI assets (default `/`, optional base path) and can host a live‑reload Canvas host for nodes (`src/canvas-host/server.ts`), injecting the A2UI postMessage bridge.
The Gateway alone delivers a great experience. All apps are optional and add extra features.
### iOS app (apps/ios)
- **Discovery + pairing**: Bonjour discovery via `BridgeDiscoveryModel` (NWBrowser). `BridgeConnectionController` auto‑connects using Keychain token or allows manual host/port.
- **Node runtime**: `BridgeSession` (actor) maintains the `NWConnection`, hello handshake, ping/pong, RPC requests, and `invoke` callbacks.
- **Canvas**: `WKWebView` with bundled Canvas scaffold + A2UI, JS eval, snapshot capture, and `clawdbot://` deep‑link interception (`ScreenController`).
- **Voice + deep links**: voice wake sends `voice.transcript` events; `clawdbot://agent` links emit `agent.request`. Voice wake triggers sync via `voicewake.get` + `voicewake.changed`.
If you plan to build/run companion apps, follow the platform runbooks below.
## Companion apps
The **macOS app is critical**: it runs the menu‑bar control plane, owns local permissions (TCC), hosts Voice Wake, exposes WebChat/debug tools, and coordinates local/remote gateway mode. Most “assistant” UX lives here.
Env vars: loaded from `.env` in the current working directory, plus a global fallback at `~/.clawdbot/.env` (aka `$CLAWDBOT_STATE_DIR/.env`) without overriding existing values.
Optional: import missing keys from your login shell env (sources your shell profile) via config or env var:
## Security model (important)
- **Default:** tools run on the host for the **main** session, so the agent has full access when it’s just you.
- **Group/channel safety:** set `agents.defaults.sandbox.mode: "non-main"` to run **non‑main sessions** (groups/channels) inside per‑session Docker sandboxes; bash then runs in Docker for those sessions.
- Set `TELEGRAM_BOT_TOKEN` or `channels.telegram.botToken` (env wins).
- Optional: set `channels.telegram.groups` (with `channels.telegram.groups."*".requireMention`); when set, it is a group allowlist (include `"*"` to allow all). Also `channels.telegram.allowFrom` or `channels.telegram.webhookUrl` as needed.
- Set `TELEGRAM_BOT_TOKEN` or `telegram.botToken` (env wins).
- Optional: set `telegram.groups` (with `telegram.groups."*".requireMention`), `telegram.allowFrom`, or `telegram.webhookUrl` as needed.
- Set `DISCORD_BOT_TOKEN` or `channels.discord.token` (env wins).
- Optional: set `commands.native`, `commands.text`, or `commands.useAccessGroups`, plus `channels.discord.dm.allowFrom`, `channels.discord.guilds`, or `channels.discord.mediaMaxMb` as needed.
<li>CLI: move gateway service commands under <code>clawdbot gateway</code>, flatten node service commands under <code>clawdbot node</code>, and add <code>gateway probe</code> for reachability. https://docs.clawd.bot/cli/gateway https://docs.clawd.bot/cli/node</li>
<li>Exec: add elevated ask/full modes, tighten allowlist gating, and render approvals tables on write. https://docs.clawd.bot/tools/elevated https://docs.clawd.bot/tools/exec-approvals</li>
<li>Exec approvals: default to local host, add gateway/node targeting + target details, support wildcard agent allowlists, and tighten allowlist parsing/safe bins. https://docs.clawd.bot/cli/approvals https://docs.clawd.bot/tools/exec-approvals</li>
<li>Heartbeat: allow explicit session keys and active hours. (#1256) Thanks @zknicker. https://docs.clawd.bot/gateway/heartbeat</li>
<li>Sessions: add per-channel idle durations via <code>sessions.channelIdleMinutes</code>. (#1353) Thanks @cash-echo-bot.</li>
<li>Nodes: run exec-style, expose PATH in status/describe, and bootstrap PATH for node-host execution. https://docs.clawd.bot/cli/node</li>
<li>Cache: add <code>cache.ttlPrune</code> mode and auth-aware defaults for cache TTL behavior.</li>
<li>Queue: add per-channel debounce overrides for auto-reply. https://docs.clawd.bot/concepts/queue</li>
<li><strong>BREAKING:</strong> Control UI now rejects insecure HTTP without device identity by default. Use HTTPS (Tailscale Serve) or set <code>gateway.controlUi.allowInsecureAuth: true</code> to allow token-only auth. https://docs.clawd.bot/web/control-ui#insecure-http</li>
<li><strong>BREAKING:</strong> Envelope and system event timestamps now default to host-local time (was UTC) so agents don’t have to constantly convert.</li>
</ul>
<h3>Fixes</h3>
<ul>
<li>NPM package: include <code>dist/sessions</code> so <code>clawdbot agent</code> resolves session helpers in npx installs.</li>
<li>Node 25: avoid unsupported directory import by targeting <code>qrcode-terminal/vendor/QRCode/*.js</code> modules.</li>
<li>Streaming/Typing/Media: keep reply tags across streamed chunks, start typing indicators at run start, and accept MEDIA paths with spaces/tilde while preferring the message tool hint for image replies.</li>
<li>Agents/Providers: drop unsigned thinking blocks for Claude models (Google Antigravity) and enforce alphanumeric tool call ids for strict providers (Mistral/OpenRouter). (#1372) Thanks @zerone0x.</li>
<li>Exec approvals: treat main as the default agent, align node/gateway allowlist prechecks, validate resolved paths, avoid allowlist resolve races, and avoid null optional params. (#1417, #1414, #1425) Thanks @czekaj.</li>
<li>Exec/Windows: resolve Windows exec paths with extensions and handle safe-bin exe names.</li>
<li>Nodes/macOS: prompt on allowlist miss for node exec approvals, persist allowlist decisions, and flatten node invoke errors. (#1394) Thanks @ngutman.</li>
<li>Gateway: prevent multiple gateways from sharing the same config/state (singleton lock), keep auto bind loopback-first with explicit tailnet binding, and improve SSH auth handling. (#1380)</li>
<li>Control UI: remove the chat stop button, keep the composer aligned to the bottom edge, stabilize session previews, and refresh the debug panel on route-driven tab changes. (#1373) Thanks @yazinsai.</li>
<li>UI/config: export <code>SECTION_META</code> for config form modules. (#1418) Thanks @MaudeBot.</li>
<li>macOS: keep chat pinned during streaming replies, include Textual resources, respect wildcard exec approvals, allow SSH agent auth, and default distribution builds to universal binaries. (#1279, #1362, #1384, #1396) Thanks @ameno-, @JustYannicc.</li>
<li>BlueBubbles: resolve short message IDs safely, expose full IDs in templates, and harden short-id fetch wrappers. (#1369, #1387) Thanks @tyler6204.</li>
<li>Models/Configure: inherit session model overrides in threads/topics, map OpenCode Zen models to the correct APIs, narrow Anthropic OAuth allowlist handling, seed allowlist fallbacks, list the full catalog when no allowlist is set, and limit <code>/model</code> list output. (#1376, #1416)</li>
<li>Memory: prevent CLI hangs by deferring vector probes, add sqlite-vec/embedding timeouts, and make session memory indexing async.</li>
<li>Cron: cap reminder context history to 10 messages and honor <code>contextMessages</code>. (#1103) Thanks @mkbehr.</li>
<li>Cache: restore the 1h cache TTL option and reset the pruning window.</li>
<li>Infra: preserve fetch helper methods/preconnect when wrapping abort signals and normalize Telegram fetch aborts.</li>
<li>Config/Doctor: avoid stack traces for invalid configs, log the config path, avoid WhatsApp config resurrection, and warn when <code>gateway.mode</code> is unset. (#900)</li>
<li>Memory: render progress immediately, color batch statuses in verbose logs, and poll OpenAI batch status every 2s by default. https://docs.clawd.bot/concepts/memory</li>
<li>Memory: add <code>--verbose</code> logging for memory status + batch indexing details. https://docs.clawd.bot/concepts/memory</li>
<li>Browser: allow config defaults for efficient snapshots in the tool/CLI. (#1336) https://docs.clawd.bot/tools/browser</li>
<li>Nostr: add the Nostr channel plugin with profile management + onboarding defaults. (#1323) https://docs.clawd.bot/channels/nostr</li>
<li>Matrix: migrate to matrix-bot-sdk with E2EE support, location handling, and group allowlist upgrades. (#1298) https://docs.clawd.bot/channels/matrix</li>
<li>Build: use tsgo for dev/watch builds by default (opt out with <code>CLAWDBOT_TS_COMPILER=tsc</code>).</li>
<li>Repo: remove the Peekaboo git submodule now that the SPM release is used.</li>
<li>macOS: switch PeekabooBridge integration to the tagged Swift Package Manager release.</li>
<li>macOS: stop syncing Peekaboo in postinstall.</li>
<li>Swabble: use the tagged Commander Swift package release.</li>
</ul>
<h3>Breaking</h3>
<ul>
<li><strong>BREAKING:</strong> Reject invalid/unknown config entries and refuse to start the gateway for safety. Run <code>clawdbot doctor --fix</code> to repair, then update plugins (<code>clawdbot plugins update</code>) if you use any.</li>
</ul>
<h3>Fixes</h3>
<ul>
<li>NPM package: include <code>dist/sessions</code> so <code>clawdbot agent</code> resolves session helpers in npx installs.</li>
<li>Node 25: avoid unsupported directory import by targeting <code>qrcode-terminal/vendor/QRCode/index.js</code>.</li>
<li>Discovery: shorten Bonjour DNS-SD service type to <code>_clawdbot-gw._tcp</code> and update discovery clients/docs.</li>
<li>Agents: propagate accountId into embedded runs so sub-agent announce routing honors the originating account. (#1058)</li>
<li>Agents: avoid treating timeout errors with "aborted" messages as user aborts, so model fallback still runs. (#1137)</li>
<li>Agents: sanitize oversized image payloads before send and surface image-dimension errors.</li>
<li>Sessions: fall back to session labels when listing display names. (#1124)</li>
<li>Compaction: include tool failure summaries in safeguard compaction to prevent retry loops. (#1084)</li>
<li>Config: log invalid config issues once per run and keep invalid-config errors stackless.</li>
<li>Config: allow Perplexity as a web_search provider in config validation. (#1230)</li>
<li>Config: allow custom fields under <code>skills.entries.<name>.config</code> for skill credentials/config. (#1226)</li>
<li>Doctor: clarify plugin auto-enable hint text in the startup banner.</li>
<li>Doctor: canonicalize legacy session keys in session stores to prevent stale metadata. (#1169)</li>
<li>Docs: make docs:list fail fast with a clear error if the docs directory is missing.</li>
<li>Plugins: add Nextcloud Talk manifest for plugin config validation. (#1297)</li>
<li>Plugins: surface plugin load/register/config errors in gateway logs with plugin/source context.</li>
<li>CLI: preserve cron delivery settings when editing message payloads. (#1322)</li>
<li>CLI: keep <code>clawdbot logs</code> output resilient to broken pipes while preserving progress output.</li>
<li>CLI: avoid duplicating --profile/--dev flags when formatting commands.</li>
<li>CLI: centralize CLI command registration to keep fast-path routing and program wiring in sync. (#1207)</li>
<li>CLI: keep banners on routed commands, restore config guarding outside fast-path routing, and tighten fast-path flag parsing while skipping console capture for extra speed. (#1195)</li>
<li>CLI: skip runner rebuilds when dist is fresh. (#1231)</li>
<li>CLI: add WSL2/systemd unavailable hints in daemon status/doctor output.</li>
<li>Status: route native <code>/status</code> to the active agent so model selection reflects the correct profile. (#1301)</li>
<li>Status: show both usage windows with reset hints when usage data is available. (#1101)</li>
<li>UI: keep config form enums typed, preserve empty strings, protect sensitive defaults, and deepen config search. (#1315)</li>
<li>UI: preserve ordered list numbering in chat markdown. (#1341)</li>
<li>UI: allow Control UI to read gatewayUrl from URL params for remote WebSocket targets. (#1342)</li>
<li>UI: prevent double-scroll in Control UI chat by locking chat layout to the viewport. (#1283)</li>
<li>UI: enable shell mode for sync Windows spawns to avoid <code>pnpm ui:build</code> EINVAL. (#1212)</li>
<li>TUI: keep thinking blocks ordered before content during streaming and isolate per-run assembly. (#1202)</li>
<li>TUI: align custom editor initialization with the latest pi-tui API. (#1298)</li>
<li>TUI: show generic empty-state text for searchable pickers. (#1201)</li>
<li>TUI: highlight model search matches and stabilize search ordering.</li>
<li>Configure: hide OpenRouter auto routing model from the model picker. (#1182)</li>
<li>Memory: show total file counts + scan issues in <code>clawdbot memory status</code>.</li>
<li>Memory: fall back to non-batch embeddings after repeated batch failures.</li>
<li>Memory: apply OpenAI batch defaults even without explicit remote config.</li>
<li>Memory: index atomically so failed reindex preserves the previous memory database. (#1151)</li>
Some files were not shown because too many files have changed in this diff
Show More
Reference in New Issue
Block a user
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.