菜单
Use when user wants to ingest a development-skills feedback report and apply fixes, or runs /ingest-feedback. Challenges every suggestion against the Iron Rules before accepting; most friction points should SKIP. Expect a report path as argument.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Temporary inboxes for agent signup QA — create disposable addresses, wait for verification email, extract OTP or magic links. Use when automating signup, login, or email verification without parsing HTML manually.
Spin up isolated sandboxes ("boxes") for coding agents, run them in parallel, queue background runs with -i, and push commits safely through the host relay. Use when the user wants to run Claude Code / Codex / OpenCode in a sandbox, start more boxes, attach to a running box, or otherwise operate the `agentbox` CLI on their laptop.
Use when building an HTTP val — a web endpoint, API route, webhook receiver, or any val that responds to HTTP requests. Covers the handler signature, Hono usage, the endpoint URL, CORS behavior, redirects, and Val Town-specific limitations.
Use when creating a new val and choosing which starter to fork. Covers the catalog of official starter templates, which project shape each one fits, and when forking an existing val is the better starting point.
Switch coding-agent accounts on a usage/rate limit, routed by host and agent.
Use Agent Mail from Codex for file leases, notifications, inboxes, and conflict prevention.
| name | ingest-feedback |
| description | Use when user wants to ingest a development-skills feedback report and apply fixes, or runs /ingest-feedback. Challenges every suggestion against the Iron Rules before accepting; most friction points should SKIP. Expect a report path as argument. |
| argument-hint | <report-path> |
| disable-model-invocation | true |
Ingest a development-skills feedback report. Challenge every suggestion against the Iron Rules before accepting. Report path: $ARGUMENTS
The report describes what happened, not what should change. Most friction is model behavior or edge cases. A change must EARN its place.
Iron Rules to apply here (canonical: ../../shared/iron-rules.md):
Marginal improvement + added words/rules → SKIP.
Extract friction points (Section 3), proposed evals (Section 4), chain-of-thought (Section 2).
Read every file referenced in friction points. Understand current instruction, why it exists, what went wrong.
Research content relevant to friction points:
@~/Documents/ai/superpowers@~/Documents/ai/claude-code-best-practiceWrite verdicts in docs/reports/ingest-YYYY-MM-DD.md.
| Verdict | Meaning | Burden |
|---|---|---|
| FIX | Instruction is wrong, misleading, or causes repeated waste. Net simplification or correction. | HIGH |
| SKIP | Edge case, model behavior, already handled, or fix adds more complexity. | LOW (default) |
| INVESTIGATE | Insufficient information. | — |
| # | Friction Point | Verdict | Rejection filter? | Reasoning |
|---|
| Verdict | Meaning |
|---|---|
| ADD | Real scenario tied to a FIX verdict |
| SKIP | Tied to SKIP, duplicates existing, or tests model behavior |
| MERGE | Combine with existing eval |
Rule: ADD only if friction point is FIX. Cross-reference evals/evals.json for duplicates.
FIX verdicts only. Surgical edits:
evals.json with next IDVerify evals.json is valid JSON. Run pre-commit run (if .py touched: ruff format . + ruff check . --fix).
Report: FIX/SKIP/INVESTIGATE counts, files changed and why, evals added/merged/skipped, report location.
Expected: most friction points SKIP. If >30% FIX, re-examine critical rigor.