agentleFS
Dimon94Dimon94/restart-life/.agents/skills/life-prove/SKILL.md

life-prove

Skill1 starsChanged 7 months agoMITView on GitHub

证据评审层。Use when an experiment log exists and the user asks whether a direction is reliable, whether to continue, or whether profile/compass/roadmap can be updated.

The file

We split this skill into the rules it gives an agent

3 rules. About 0 are common, 0 are distinctive, and 3 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈0

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 3

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Read references/guide.md for 思想来源/source lineage, proof verdict, and evidence separation rules.

    line 109
  2. Very specific:

    Use assets/proof-card.template.json as the proof-card template material.

    line 110
  3. Very specific:

    Run node scripts/check-handoff.mjs --runtime ~/.restart-life --cycle <cycle> before handing off to life-integrate.

    line 111

≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

life-prove

你负责判断实验是否产生了足够证据。你的结论只能是 pass、fail 或 blocked,不能把一次情绪波动包装成长期人生结论。

铁律

  • NO COMPASS UPDATE WITHOUT EVIDENCE: 没有证据,不更新 compass/profile/roadmap。
  • synthetic E2 evidence cannot update compass.
  • mood 是信号,不是判决;必须和实际行动、外部反馈、产出证据分开。

入口检查

  1. 读取 experiment-plan.md,确认原 hypothesis、evidence to collect、exit signal。
  2. 读取 experiment-log.jsonl 和 evidence-ledger.jsonl,核对 evidence ids。
  3. 若实验没有执行或证据太少,直接 blocked,reroute 到 life-act。
  4. 若证据反驳了原问题框架,reroute 到 life-investigate,不要硬写结论。

Proof Card

写入 ~/.restart-life/cycles/<cycle>/proof-card.json:

{
  "cycle_id": "RL-001-career-direction",
  "verdict": "pass|fail|blocked",
  "hypothesis": "...",
  "evidence_ids": ["E-101", "E-102"],
  "supporting_signals": ["..."],
  "contradicting_signals": ["..."],
  "mood_signals": ["..."],
  "confidence": "low|medium|high",
  "durable_update_allowed": false,
  "recommended_next_route": "life-integrate|life-act|life-investigate"
}

durable_update_allowed 只能在非 synthetic evidence 足够、且 verdict 为 pass 或明确 fail 学习时为 true。blocked 永远不能写 durable update。

判断标准

Verdict使用条件下一步
pass证据支持继续或升级方向life-integrate
fail证据反驳假设,且学到了可记录内容life-integrate 或 life-investigate
blocked实验未执行、证据不足、记录不可信life-act

执行流程

  1. 对照 experiment plan,列出已收集证据和缺失证据。
  2. 分离 action/output/feedback/energy/friction,不混成单一感受。
  3. 判断支持信号和反对信号。
  4. 给出 verdict、confidence、durable_update_allowed。
  5. handoff 到下一 route,不直接改 durable artifacts。

质量门

交给 life-integrate 前运行:

npm run validate:life-cycle -- --runtime ~/.restart-life --cycle <cycle> --mode <mode>

禁止事项

  • 不因为用户“感觉不错”就 pass。
  • 不用 roundtable 或 AI 观点作为 compass 更新证据。
  • 不在 proof 阶段补写实验记录;缺什么就判 blocked。
  • 不把 proof-card.json 写成安慰性总结。

flows in skill-manifest.json declares this skill's next route to life-integrate when proof is usable.

Bundled Resources

  • Read references/guide.md for 思想来源/source lineage, proof verdict, and evidence separation rules.
  • Use assets/proof-card.template.json as the proof-card template material.
  • Run node scripts/check-handoff.mjs --runtime ~/.restart-life --cycle <cycle> before handing off to life-integrate.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.