القائمة
Cross-session learnings lifecycle — schema, storage, retrieval, and pruning of lessons learned during orchestrator sessions. Use when managing learnings via the /learn command.
التثبيت باستخدام Codex أو Claude انسخ هذا Prompt والصقه في Codex أو Claude أو مساعد آخر ليراجع صفحة Skill ويثبّتها لك.
استنادا إلى تصنيف SOC المهني
Demo pipeline state machine — 7-phase autonomous sequence with delegation context templates, phase transition logic, BLOCKED recovery strategies, and demo-state.json schema. Used exclusively by demo-conductor.
Cinematic narration style guide for demo-conductor — ANSI-coloured banner formats, live pipeline scoreboard, audience-facing language, phase summaries, and error narration patterns. Keeps the demo presentation-quality throughout.
Standardized completion and escalation protocol for subagent responses. Ensures the conductor can machine-parse every subagent return. Use when reporting completion status back to the orchestrator.
Agent Teams assembly and task injection — selects appropriate team, validates prerequisites, estimates cost, injects tasks into the shared task list, and manages team lifecycle.
Token and cost tracking with model tier enforcement
Complexity-based task routing with model tier selection
| name | learnings-mgmt |
| description | Cross-session learnings lifecycle — schema, storage, retrieval, and pruning of lessons learned during orchestrator sessions. Use when managing learnings via the /learn command. |
| user-invocable | false |
Learnings persist to artifacts/memory/learnings.jsonl — one JSON object per line.
{
"id": "learn-<timestamp-ms>",
"timestamp": "2026-03-30T12:00:00.000Z",
"category": "pattern | pitfall | convention | decision | tooling",
"note": "Free-text lesson learned",
"source": "session | user | review | incident",
"deleted": false
}
| Field | Required | Description |
|---|---|---|
id | Yes | learn- prefix + Unix epoch milliseconds — unique per entry |
timestamp | Yes | ISO 8601 |
category | Yes | One of: pattern, pitfall, convention, decision, tooling |
note | Yes | The lesson text — keep under 200 words |
source | Yes | One of: session, user, review, incident |
deleted | No | Defaults to false. Set true on soft-delete. |
Append a new JSON line to learnings.jsonl. Auto-assign id and timestamp. Default source to session unless the user specifies otherwise.
Read learnings.jsonl, filter out deleted: true entries, display the last 20 in reverse chronological order. Show: id, timestamp (date only), category, and truncated note (first 80 chars).
Read all non-deleted entries, filter where note contains the query string (case-insensitive). Display matching entries with full note text.
Find the entry by id. Set deleted: true and rewrite the line in place. Do not physically delete — this preserves audit history.
Read all non-deleted entries. Write a Markdown summary to artifacts/memory/learnings-export.md grouped by category. Include timestamp and full note text.
Agents may suggest adding a learning when:
The agent should present the suggestion and let the user confirm via /learn add <note>.
Over time, learnings accumulate. Recommend pruning when learnings.jsonl exceeds 200 entries:
artifacts/memory/learnings-archive.jsonl