Commit b3636b1
Release v4.7.0: Event-Driven Team Runtime & Multi-Model Flexibility (#1354)
* Fix HUD build guidance and detection for plugin cache (#1328) (#1332)
* feat(hud): add wrap mode for maxWidth output (#1331)
* fix: use $argv when spawning workers under fish shell (#1329)
* fix(team): use $argv for fish worker launch forwarding
* docs(changelog): note fish worker spawn argv fix
* fix(hud): resolve documentation and error handling slop (#1307)
* chore(release): bump version to 4.6.1
* fix(hud): resolve documentation and error handling slop
Fix SKILL.md documentation to match implementation:
- Remove non-existent showCache and showCost properties
- Add proper safeMode documentation
- Update config examples to match HudElementConfig type
- Clarify CLAUDE_CONFIG_DIR support for config paths
- Add agentsFormat options documentation
Fix error handling in state.ts and index.ts:
- Add error logging to all catch blocks
- writeHudState now logs failures instead of silent return
- readHudState logs read/parse errors
- Config operations log errors to stderr
- Debug-guarded error logging in index.ts
All 281 HUD tests pass.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
* fix(benchmark): harden harsh-critic parser for markdown variants (#1301)
Rebased onto dev modular scoring structure. Parser changes already
present on dev; this commit adds the test suite covering:
- Markdown heading sections and bold-number findings
- Perspective subsection headings
- None gap analysis edge case
- function():line-range evidence markers
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
* feat(benchmark): calibrate keyword matcher with deterministic thresholds (#1300)
Rebased onto dev modular scoring structure. Scorer changes already
present on dev; this commit adds:
- Test suite for punctuation/hyphen robustness
- Dynamic threshold boundary tests (6-keyword negative/positive)
- Scoring calibration rationale document
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
* docs: remove deprecated analytics CLI commands (#1336) (#1341)
* feat(notifications): add custom integration system with webhook and CLI support
Add generic extensibility to notification system via custom webhooks and
CLI commands, consolidating configure-openclaw into configure-notifications.
New Features:
- Custom webhook integrations with template variable support
- Custom CLI command integrations (execFile, not shell)
- Built-in presets: OpenClaw, n8n, ClawdBot, generic webhook/CLI
- Automatic migration from legacy OpenClaw config
- Interactive wizard in configure-notifications skill
Security:
- HTTPS enforcement for webhooks (except localhost)
- Timeout bounds (1-60s)
- CRLF injection prevention in headers
- Shell metacharacter rejection in CLI args
- execFile (not exec) for CLI commands
Files Added:
- src/notifications/template-variables.ts
- src/notifications/validation.ts
- src/notifications/presets.ts
- src/notifications/__tests__/custom-integration.test.ts
Files Modified:
- src/notifications/types.ts (CustomIntegration types)
- src/notifications/config.ts (config reader + migration)
- src/notifications/dispatcher.ts (webhook/CLI dispatch)
- src/notifications/index.ts (exports)
- src/hooks/session-end/index.ts (remove OpenClaw code)
- skills/configure-notifications/SKILL.md (custom wizard)
- skills/configure-openclaw/SKILL.md (deprecation notice)
Tests: 31 new tests, all 522 notification tests passing
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* test(qa): add integration tests for custom webhook and CLI dispatch
Add end-to-end QA tests that verify:
- Webhook dispatch with template interpolation
- CLI command execution via execFile
- Custom header handling
Run with: node scripts/qa-tests/test-custom-integration.mjs
* test: remove placeholder example test suite
* test: add runtime and integration edge-case coverage
* fix(cli): handle bare --notify without consuming next arg
* test: fix stop-hook OpenClaw cooldown assertion (#1344)
* test: fix stop-hook OpenClaw cooldown assertion
* test: align session-end openclaw test with current hook behavior
* fix(session-end): restore awaited OpenClaw wake on session end
* Revert "fix(session-end): restore awaited OpenClaw wake on session end"
This reverts commit 822e54f.
* feat(cli): integrate omx ask/team into omc; deprecate team MCP runtime
* feat(team): event-driven team redesign matching OMX architecture
Implement event-driven team runtime (v2) with identical CLI API surface
to oh-my-codex. Key changes:
- Add 28-operation CLI API (api-interop.ts) identical to OMX
- Add team-ops.ts MCP gateway with 40+ functions
- Add runtime-v2.ts with snapshot-based monitoring (no polling watchdog)
- Add dispatch queue pattern with file-based O_EXCL locking
- Add worker/dispatch/leader-nudge hooks for event-driven coordination
- Add dynamic scaling (scale up/down) gated by OMC_TEAM_SCALING_ENABLED
- Add JSONL append-only event log for observability
- Add task claim/transition/release with 15-min lease
- Remove done.json entirely - completion via CLI API lifecycle transitions
- Fix OMX_TEAM_AUTO_INTERRUPT_RETRY naming leak to OMC_ prefix
- Feature-flagged behind OMC_RUNTIME_V2=1
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* feat(team): v2 runtime uses direct tmux spawn with CLI API inbox, no done.json or watchdog
- Rewrite startTeamV2() to stop delegating to v1's startTeam()
- Add buildV2TaskInstruction() with CLI API lifecycle commands
- Add spawnV2Worker() for direct tmux pane creation
- Fix runtime-cli.ts to use startTeamV2/shutdownTeamV2 when OMC_RUNTIME_V2=1
- Add omc team CLI entry point (start/status/shutdown/api subcommands)
- Rebuild bridge bundles with v2 code paths
- Verified: claude and codex workers spawn with CLI API inbox, no done.json
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* fix: stop infinite 401/403 OAuth loops in Team persistent mode (#1308) (#1330)
* fix(team): stop auth-error continuation loops in persistent mode
* test: mock isAuthenticationError in stop hook cooldown test
* fix(cli): remove duplicate 'team' command registration
The legacy team command (runTeamCommand + TEAM_COMMAND_USAGE) was not
removed when the new commands/team.ts handler was integrated, causing
Commander to throw "cannot add command 'team' as already have command
'team'" on every CLI invocation.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* test(cli): add CLI boot regression tests for duplicate command registration
Adds static analysis to detect duplicate .command() names in the CLI
source, plus runtime tests ensuring omc --help, --version, and --madmax
boot without crashing.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* feat(ask): add ask-codex and ask-gemini skills (#1350)
* feat(ask): add ask-codex and ask-gemini skills with provider-specific CLI args
- Add codex as valid ask provider alongside claude and gemini
- Create ask-codex and ask-gemini skill surfaces (SKILL.md) for plugin discovery
- Add wrapper scripts (scripts/ask-codex.sh, scripts/ask-gemini.sh) with
PATH-independent repo-local CLI entrypoint execution
- Use provider-specific CLI args: codex exec --dangerously-bypass-approvals-and-sandbox,
gemini -p --yolo, claude -p
- Fix test env: strip CLAUDECODE env var from spawned test processes to prevent
nested session detection failures
- Update skill tests with alias-aware counts (38/37/38) and ask-codex expectations
- Add wrapper wiring tests: provider dispatch, prompt forwarding, env passthrough,
exit code propagation, PATH-shadow isolation
- Update docs (REFERENCE.md, CLAUDE.md, AGENTS.md) with ask-codex references
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* chore: rebuild bridge/cli.cjs with codex provider support
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* fix: resolve pre-existing lint errors and team-help test failure
- Fix empty catch blocks in runtime-cli.ts, runtime-v2.ts, scaling.ts
- Replace empty interface with type alias in state/tasks.ts
- Update team-help test to check team.ts instead of index.ts
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
* fix(skills): use CLAUDE_PLUGIN_ROOT for ask skill wrapper paths (#1351)
Wrapper paths in SKILL.md used relative ./scripts/ which doesn't
resolve in production where the plugin is installed at a different
location. Use ${CLAUDE_PLUGIN_ROOT}/scripts/ consistent with other
skills like team and cancel.
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
* docs(skills): align team/ask docs and route ccg via ask skills
* chore: bump version to 4.7.0 and add release notes
Event-driven team runtime v2, ask-codex/ask-gemini skills for non-tmux
multi-model flexibility, 15 security fixes, custom notifications, and
HUD enhancements. 50+ PRs merged, 30K+ lines changed.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
Co-authored-by: Alex Urevick-Ackelsberg <alex@zivtech.com>1 parent d45a323 commit b3636b1
File tree
342 files changed
+34620
-2609
lines changed- .claude-plugin
- benchmarks/harsh-critic
- scoring/__tests__
- bridge
- dist
- __tests__
- hud
- rate-limit-wait
- cli
- __tests__
- commands
- __tests__
- hooks
- __tests__
- persistent-mode
- __tests__
- session-end
- __tests__
- todo-continuation
- __tests__
- hud
- installer
- __tests__
- mcp
- __tests__
- notifications
- __tests__
- team
- __tests__
- state
- docs
- ko
- scripts
- qa-tests
- skills
- ask-codex
- ask-gemini
- ccg
- configure-notifications
- configure-openclaw
- hud
- omc-setup
- omc-teams
- team
- src
- __tests__
- hud
- cli
- __tests__
- commands
- __tests__
- hooks
- __tests__
- persistent-mode
- __tests__
- session-end
- __tests__
- todo-continuation
- __tests__
- hud
- installer
- mcp
- __tests__
- notifications
- __tests__
- team
- __tests__
- state
- templates/hooks
Some content is hidden
Large Commits have some content hidden by default. Use the searchbox below for content that may be hidden.
342 files changed
+34620
-2609
lines changed| Original file line number | Diff line number | Diff line change | |
|---|---|---|---|
| |||
10 | 10 | | |
11 | 11 | | |
12 | 12 | | |
13 | | - | |
| 13 | + | |
14 | 14 | | |
15 | 15 | | |
16 | 16 | | |
| |||
27 | 27 | | |
28 | 28 | | |
29 | 29 | | |
30 | | - | |
| 30 | + | |
31 | 31 | | |
| Original file line number | Diff line number | Diff line change | |
|---|---|---|---|
| |||
1 | 1 | | |
2 | 2 | | |
3 | | - | |
| 3 | + | |
4 | 4 | | |
5 | 5 | | |
6 | 6 | | |
| |||
| Original file line number | Diff line number | Diff line change | |
|---|---|---|---|
| |||
| 1 | + | |
| 2 | + | |
| 3 | + | |
| 4 | + | |
| 5 | + | |
| 6 | + | |
| 7 | + | |
| 8 | + | |
| 9 | + | |
| 10 | + | |
| 11 | + | |
| 12 | + | |
| 13 | + | |
| 14 | + | |
| 15 | + | |
| 16 | + | |
| 17 | + | |
| 18 | + | |
| 19 | + | |
| 20 | + | |
| 21 | + | |
| 22 | + | |
| 23 | + | |
| 24 | + | |
| 25 | + | |
| 26 | + | |
| 27 | + | |
| 28 | + | |
| 29 | + | |
| 30 | + | |
| 31 | + | |
| 32 | + | |
| 33 | + | |
| 34 | + | |
| 35 | + | |
| 36 | + | |
| 37 | + | |
| 38 | + | |
| 39 | + | |
| 40 | + | |
| 41 | + | |
| 42 | + | |
| 43 | + | |
| 44 | + | |
| 45 | + | |
| 46 | + | |
| 47 | + | |
| 48 | + | |
| 49 | + | |
| 50 | + | |
| 51 | + | |
| 52 | + | |
| 53 | + | |
| 54 | + | |
| 55 | + | |
| 56 | + | |
| 57 | + | |
| 58 | + | |
| 59 | + | |
| 60 | + | |
| 61 | + | |
| 62 | + | |
| 63 | + | |
| 64 | + | |
| 65 | + | |
| 66 | + | |
| 67 | + | |
| 68 | + | |
| 69 | + | |
| 70 | + | |
| 71 | + | |
| 72 | + | |
| 73 | + | |
| 74 | + | |
| 75 | + | |
| 76 | + | |
| 77 | + | |
| 78 | + | |
| 79 | + | |
| 80 | + | |
| 81 | + | |
| 82 | + | |
| 83 | + | |
| 84 | + | |
| 85 | + | |
| 86 | + | |
| 87 | + | |
| 88 | + | |
| 89 | + | |
| 90 | + | |
| 91 | + | |
| 92 | + | |
| 93 | + | |
| 94 | + | |
| 95 | + | |
| 96 | + | |
| 97 | + | |
1 | 98 | | |
2 | 99 | | |
3 | 100 | | |
| |||
20 | 117 | | |
21 | 118 | | |
22 | 119 | | |
| 120 | + | |
| 121 | + | |
| 122 | + | |
| 123 | + | |
| 124 | + | |
| 125 | + | |
23 | 126 | | |
24 | 127 | | |
25 | 128 | | |
| |||
0 commit comments