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。
直接命令不会经过审查 Prompt;运行前请先检查来源。
npx skills add https://github.com/benjam3n/reasoningtool --skill sacri命令会保持在同一行。复制前请横向滚动并检查完整内容。
想先保存到本地?可下载 SkillsMP 当前能够提供的文件。
Route any input through a branching question tree to narrow down the optimal response strategy before writing. Two stages — PERCEIVE (classify input) then ACT (select response). Covers all prompt types.
Generate exhaustive guesses about user input using ALL search methods with coverage tracking. Guessing is SEARCH through possibility space. Tracks space created vs space covered to ensure comprehensive exploration.
Systematically evaluate and select from a set of guesses, options, or possibilities. Combines ARAW analysis with prioritization to determine which guesses are strong, weak, actionable, or eliminable.
基于 SOC 职业分类
正在显示 SKILL.md
| name | sacri |
| description | Ensure gates and procedures link to other procedures that provide the needed answer-interface and next steps. |
| output | {"format":"prose"} |
Input: $ARGUMENTS
Ensure gates and procedures link to other procedures that provide the needed answer-interface and next steps. Broken or missing cross-references mean executors hit dead ends.
From the target artifact, extract:
For each reference:
| Reference | Target | Exists? | Accessible? |
|---|---|---|---|
| [reference text] | [what it points to] | Y/N | Y/N |
If a reference points to a non-existent procedure: flag as BROKEN.
For each existing reference, check: does the referenced procedure provide what the referrer needs?
| Referrer Needs | Referenced Provides | Match? |
|---|---|---|
| Definition (what does X mean?) | [what it actually provides] | Y/N |
| Observation (what to look for?) | [what it actually provides] | Y/N |
| Measurement (what metric?) | [what it actually provides] | Y/N |
| Provenance (what source?) | [what it actually provides] | Y/N |
| Decision (what choice?) | [what it actually provides] | Y/N |
If the referenced procedure doesn't provide what the referrer needs: flag as MISMATCHED.
Check for places where a cross-reference SHOULD exist but doesn't:
For each issue:
| Issue | Type | Fix |
|---|---|---|
| [reference] | broken | [correct target or remove] |
| [reference] | mismatched | [better target or note the gap] |
| [location] | missing | [add reference to: procedure_name] |
CROSS-REFERENCE INTEGRITY:
Target: [what was audited]
References found: [N]
Intact: [N]
Broken: [N]
Mismatched: [N]
Missing references: [N]
Issues:
| # | Location | Issue | Severity | Fix |
|---|----------|-------|----------|-----|
| 1 | [where] | [what's wrong] | [H/M/L] | [how to fix] |
Integrity score: [intact / (intact + broken + mismatched)]%