claude-teammate 0.1.285 → 0.1.286
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/README.md +31 -0
- package/package.json +1 -1
package/README.md
CHANGED
|
@@ -44,6 +44,7 @@ Assign it tickets. It ships.
|
|
|
44
44
|
- **Stuck-task auto-retry** - if Claude CLI fails mid-task, the worker automatically retries up to 5 times before escalating with a Jira comment.
|
|
45
45
|
- **Live dashboard** - real-time heartbeat cards, stuck-task table, and per-issue log viewer with download.
|
|
46
46
|
- **No-code deliverable attachments** - for research and document tasks the bot uploads output files directly to the Jira issue.
|
|
47
|
+
- **Self-improving skills** - the worker auto-detects failed/bypassed Skill tool calls and opens PRs that patch the offending skill file (Phase 4 also samples clean tasks for proactive improvements).
|
|
47
48
|
|
|
48
49
|
<br/>
|
|
49
50
|
|
|
@@ -187,6 +188,7 @@ Runtime files and memory are stored in:
|
|
|
187
188
|
├── worker.pid
|
|
188
189
|
└── state.json
|
|
189
190
|
└── memory/
|
|
191
|
+
├── skill-fixes.json # self-improve event log (rolling, max 200)
|
|
190
192
|
└── {domain}/
|
|
191
193
|
└── {workspace}/
|
|
192
194
|
├── epic-{jira-key}.md
|
|
@@ -195,6 +197,35 @@ Runtime files and memory are stored in:
|
|
|
195
197
|
|
|
196
198
|
<br/>
|
|
197
199
|
|
|
200
|
+
## Self-Improving Skills
|
|
201
|
+
|
|
202
|
+
The worker watches every Claude task for Skill tool failures (direct errors, MUST/MANDATORY bypass, silent fallback to a different tool). When detected, it auto-generates a patch and opens a PR against the skill file — repo-scoped skills get a PR in that repo, global skills (`~/.claude/skills/<name>`) are patched in place with a timestamped backup under `.backups/`.
|
|
203
|
+
|
|
204
|
+
**Always-on (reactive fix mode)** — no setup required. Cooldown: 30 min, threshold 2 successful fixes per `(skill, errorType)`.
|
|
205
|
+
|
|
206
|
+
**Optional: Phase 4 proactive improvement** — on otherwise-clean tasks, sample a fraction and ask Claude whether the skill could be improved. Disabled by default (cost-bounded). To enable, add to `~/.tm8/.env`:
|
|
207
|
+
|
|
208
|
+
```env
|
|
209
|
+
SKILL_IMPROVEMENT_PROACTIVE=true
|
|
210
|
+
SKILL_IMPROVEMENT_SAMPLE_RATE=0.1 # fraction of clean tasks sampled (0.0–1.0)
|
|
211
|
+
```
|
|
212
|
+
|
|
213
|
+
Restart the worker (`tm8 stop && tm8 start`) to pick up the change. Cooldown: 24h per skill.
|
|
214
|
+
|
|
215
|
+
**Optional override**: `TM8_EVENTS_ROOT=/custom/path` redirects `memory/skill-fixes.json` away from `~/.tm8` (tests, multi-tenant deploys).
|
|
216
|
+
|
|
217
|
+
**Inspect activity:**
|
|
218
|
+
```bash
|
|
219
|
+
cat ~/.tm8/memory/skill-fixes.json # rolling log of fix attempts + statuses
|
|
220
|
+
tm8 dashboard # /api/skill-fixes + /skill-fixes/stats UI
|
|
221
|
+
```
|
|
222
|
+
|
|
223
|
+
Event statuses: `patched`, `patched-with-backup`, `pr-created`, `no-fix`, `cooldown`, `lock-skipped`, `not-found`, `scope-excluded`, `patch-failed`, `generation-error`.
|
|
224
|
+
|
|
225
|
+
Detection is fire-and-forget — never blocks or affects task return values.
|
|
226
|
+
|
|
227
|
+
<br/>
|
|
228
|
+
|
|
198
229
|
## Google Chat Integration
|
|
199
230
|
|
|
200
231
|
Claude Teammate can post ticket updates and receive replies via a Google Chat bot.
|