메뉴
Use when the user wants to launch a runbook, run an audit, review, or analysis task via Jetty, or says "run the X runbook". Also triggers on "launch runbook", "execute runbook", "run dedup-audit", "run security-reviewer", or any request to run a task from runbooks/.
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
SOC 직업 분류 기준
Consolidate all /review-capture drop files for the current branch into a single verified, deduplicated, conflict-annotated merged plan. Verifies each finding against HEAD (discarding stale ones), merges duplicates across reviewers, flags conflicting suggested fixes, groups findings into execution waves by file ownership, and deletes the individual drop files on success. Run after all parallel review terminals have completed.
Wrapper skill that invokes any review skill/command and captures its findings into a structured YAML drop file under .claude/review-drops/<branch>/. Use when running parallel code reviews across multiple terminals (each terminal captures one reviewer's output) so a later /merge-reviews pass can dedup, verify, and consolidate before execution. Accepts any target review skill (e.g. ce:review, multi-reviewer-patterns, asm-forge, review-pipeline, perf-pipeline, cache-correctness-review, security-reviewer, etc.).
Use when creating any beads task — auto-researches the codebase, links related tasks, and produces a rich self-contained description from a structured template. Accepts minimal intent and outputs a complete task ready for agent implementation.
Use when you want review AND automated fixes in one pass, when /review-dispatch alone would leave findings unaddressed, or before merging a feature branch that needs thorough diagnosis and remediation. Two-phase diagnose-then-fix pipeline.
Use when a beads task exists and needs validation before implementation — verifies codebase references, identifies edge cases and design flaws, assesses scope and feasibility, splits oversized tasks, dispatches domain-specific skills (test-strategy, unsafe-review, dist-sys-auditor, simd-optimize, asm-forge, performance-analyzer, security-reviewer, interface-design-review, sim-review, safe-over-unsafe) for specialized enrichment, and dispatches /deep-research or /deeper-research for ambiguous areas. The complement of /create-task — ensures tasks are buttoned up and ready for mechanical implementation.
Use when creating implementation-ready beads tasks that need testing strategy, optimal implementation approach, and documentation requirements baked in — composes /create-task with parallel enrichment agents that analyze the codebase and produce concrete test specifications, algorithm/data-structure guidance, and doc quality standards so implementing agents don't need to re-research
| name | run-runbook |
| description | Use when the user wants to launch a runbook, run an audit, review, or analysis task via Jetty, or says "run the X runbook". Also triggers on "launch runbook", "execute runbook", "run dedup-audit", "run security-reviewer", or any request to run a task from runbooks/. |
Launch runbooks from runbooks/ via Jetty using run-runbook.sh.
runbook-params.json with any parameter overrides the user requested./run-runbook.sh <task-name>If runbook-params.json is missing, create it locally at repo root before
launching. The repo .gitignore excludes both runbook-params.json and
run-runbook.sh, so these runbook helper files may exist only in your local
checkout.
NEVER write secrets or tokens into runbook-params.json. The run-runbook.sh
script injects GITHUB_TOKEN automatically via gh auth token at launch time.
Do not add a GITHUB_TOKEN field to the JSON file — it is unnecessary and
risks leaking credentials.
{
"repository": "ahrav/gossip-rs",
"pr_number": "312",
"base_branch": "main",
"CARGO_HOME": "/Users/<you>/.cargo",
"timeout_sec": 3600
}
| Parameter | Default | Override for |
|---|---|---|
timeout_sec | 3600 (1 hr) | Long audits: 7200 (2 hr), 10800 (3 hr) |
pr_number | varies | PR-specific runbooks (reviews, comment response) |
base_branch | main | Auditing a different branch |
Use the Edit tool to update runbook-params.json before running the script.
Restore the original value immediately after launching if the override was
one-time — the runbook runs remotely and does not re-read the local file.
./run-runbook.sh <task-name>
The <task-name> matches the filename in runbooks/ without the .md extension.
if [ -d runbooks ] && ls runbooks/*.md >/dev/null 2>&1; then
ls runbooks/*.md | sed 's|runbooks/||;s|\.md$||'
else
echo "No local runbooks found. Ensure local runbook assets are provisioned."
fi
The script prints trajectory and workflow IDs on success:
Run launched successfully:
Task: <task-name>
Trajectory ID: <id>
Workflow ID: <id>
Report these to the user so they can track progress in Jetty.
| Mistake | Fix |
|---|---|
| Using RemoteTrigger API directly | Use run-runbook.sh — it handles param injection and Jetty API |
| Forgetting to update timeout for long tasks | User-requested timeout goes in runbook-params.json timeout_sec |
| Passing wrong task name | Name must match a file in runbooks/ without .md. The script lists available names on error. |
| Not reporting trajectory ID | Always show the trajectory and workflow IDs in your response |
| Not knowing which runbook to run | Use the guarded runbook-listing command above and ask the user to pick one |
Writing GITHUB_TOKEN into runbook-params.json | Never. run-runbook.sh injects the token via gh auth token automatically. Writing tokens into files risks credential leaks. |