メニュー
Sync vault knowledge patterns to skills. Scans Second Brain for strong signals, proposes skill updates. Use when user says knowledge sync, sync skills, update skills from vault, or promote patterns.
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
SOC 職業分類に基づく
Execution trace capture and token baselining. Use when user says trace, extract trace, count tokens, token baseline, or /harness-trace. Not for harness optimization (use harness-mechanic).
Release prep and tagging with pre-flight verification. Use when user says release, tag, ship it, push to production, create release, or bump version. Not for editing a version string outside a release, a lockfile or a changelog entry by hand, and not for reading what changed in a past release: those are ordinary edits, just do them.
Mine LEARNING.md retrospectives across all repos for recurring failure-modes, then propose harness changes (hooks, rules, skills, checklist items) with falsifiable change-contracts. Also runs the doc-gardening pass over the governance docs (stale rules, dead cross-references, skill-table drift). Use when user says learning loop, mine learnings, recurring failures, what should I fix in my process, cross-repo retrospective, doc gardening, stale docs, or run learning-loop. Runs on human schedule, never autonomously. Not for single-project retrospectives (use learning-docs) or trace/token analysis (use harness-mechanic).
Full contractor-mode loop: sub-protocols (LOCALIZE, BENCH-BASELINE, REPRODUCE, DRIFT), review routing by file pattern, blast radius, UAT, parallelism and effort caps, escalation, just-do-it, goal-backed runs. Load it to RUN that loop, once you have an approved plan and are entering step 1, or when a step needs its detail. It is a protocol, not a way to do the work: it never replaces the language, review or planning skill the task itself calls for, and a task small enough to just edit (SKIP_SET) needs none of it. The always-on spine lives in rules/orchestrator-protocol.md. Not for trace capture (harness-trace) or harness optimization (harness-mechanic).
Ruby on Rails with service-oriented architecture, Dry-validation, Sidekiq/Solid Queue, Hotwire. Use for Rails API, Rails services, Rails forms, RSpec, ActiveRecord, Rails migrations. Not for standalone Ruby gems (use ruby skill), and not for judging whether a migration is safe to run in production (that is the database-reviewer agent).
Get three independent second opinions (isolated Claude + isolated Gemini + isolated DeepSeek) on a problem Claude is analyzing. Use when user says second opinion, another perspective, challenge this approach, or asks gemini/deepseek about a decision, approach, or diagnosis. Also invoke on your own initiative, without being asked: on a complex task before planning, after 2 failed root-cause attempts, on a 🔴 decision (rewrite, core logic, security), or when the review loop is stuck below 80 after 2 fix rounds. Gathers context, writes a focused prompt, calls all reviewers in isolated Docker containers, and synthesizes all viewpoints. Not for reviewing code or changes, even phrased as 'ask gemini to review' (use gemini-review).
| name | knowledge-sync |
| description | Sync vault knowledge patterns to skills. Scans Second Brain for strong signals, proposes skill updates. Use when user says knowledge sync, sync skills, update skills from vault, or promote patterns. |
| compatibility | Requires Obsidian CLI (obsidian MCP commands). |
Promotes recurring patterns from vault Second Brain notes into skill files. The APPLY step runs on human approval only, never autonomously. The scan-and-propose phase MAY run on a schedule (see Scheduled Mode).
| Signal | Criteria | Action |
|---|---|---|
| weak | 1 project or 1 source | Track, don't propose |
| strong | 3+ projects OR 2+ independent sources | Propose to human |
| applied | Already synced to skill | Skip |
Second Brain notes use a ## Skill Candidates table:
## Skill Candidates
<!-- Updated by learning-docs during retrospectives -->
| Pattern | Target Skill | Seen In | First Seen | Signal |
|---------|-------------|---------|------------|--------|
| Go error wrapping with %w | golang | feed-brain, bbp | 2026-01-15 | weak |
| Token bucket rate limiting | cloud-infrastructure | feed-brain, nur, clawdbot | 2026-01-10 | strong |
Table-based: easy to append via CLI, readable in Obsidian, parseable by Claude.
obsidian search query="## Skill Candidates" path="Second Brain" matchesstrong signals; skip rows where signal = applied or pattern already exists in target skillAfter human approval for each candidate:
<!-- vault: [[Second Brain - Development#Pattern Name]] -->strong to applied# Example: update signal in vault
obsidian read file="Second Brain - Development"
# Find the row, replace "strong" with "applied" in the table
obsidian append file="Second Brain - Development" content="..."
## Knowledge Sync Report
### Candidates Found: N strong, M weak (skipped)
#### golang (2 candidates)
**1. Go error wrapping with %w**
- Source: [[Second Brain - Development#Go Error Handling]]
- Seen in: feed-brain, bbp
- Proposed addition to `skills/golang/SKILL.md`:
[exact text to add]
- [ ] Approve [ ] Skip [ ] Modify
#### cloud-infrastructure (1 candidate)
...
The scan half of this skill may run unattended; the APPLY gate never does. A scheduled run executes SCAN → FILTER → GROUP → PROPOSE, writes the report to quality_reports/knowledge_sync/report-YYYY-MM-DD.md in the forge repo, and STOPS. Max reviews the report later and triggers APPLY interactively.
Runner: a user crontab entry (1st of the month, 09:23 local). The human installs it manually: persisting a headless autonomous run is a user-authorized action, Claude only documents the line (crontab -l to inspect, crontab -r or edit to remove). Replace <forge-repo> and <claude-bin> with your paths (which claude), and mkdir -p ~/.claude/logs first (cron fails silently on a missing redirect target):
23 9 1 * * cd <forge-repo> && <claude-bin> -p "Run the knowledge-sync skill in Scheduled Mode: SCAN, FILTER, GROUP, PROPOSE only. Write the report to quality_reports/knowledge_sync/report-$(date +\%Y-\%m-\%d).md. Do NOT apply any skill changes." >> ~/.claude/logs/knowledge-sync-cron.log 2>&1
Caveats:
obsidian CLI; a permission denial shows up in the log file, not on screen.