auditor
Skill0 starsChanged 4 months agoMITView on GitHub
系统优化审计 + 财务合规审计。强触发词:「审计」「优化建议」「变更审查」「系统诊断」「合规检查」「内部控制」「财务合规」。不适用:单文件格式检查、简单提问、非变更类操作。
5 other projects use a version of this.
Adopted by
Most recently taken up by
- NotRedFox/NotRedFoxs-Claude-skillstheir own version, first seen Oct 9, 2026
- opentensor/subtensortheir own version, first seen Oct 8, 2026
- QinghongLin/data2story-skilltheir own version, first seen Jul 2, 2026
- daemon-blockint-tech/Agentic-Enteprises-Skilltheir own version, first seen May 20, 2026
- rade-nugroho/Agentic-Enteprises-Skilltheir own version, first seen May 20, 2026
Other versions5 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 5
- opentensor/subtensor★ 402same name
- QinghongLin/data2story-skill★ 155same name
- rade-nugroho/Agentic-Enteprises-Skill★ 9same name
- daemon-blockint-tech/Agentic-Enteprises-Skill★ 8same name
- NotRedFox/NotRedFoxs-Claude-skills★ 1same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository5 other files it gives its agents.
- agent-methodologySkill
- agent-plannerSkill
- agent-team-orchestrationSkill
- debugSkill
- qclaw-skill-creatorSkill
The file
Files
需求路由
def route_demand(demand: str) -> str:
d = demand.lower()
if any(kw in d for kw in ["质量检查", "q0", "版本锁定"]): return "local"
if any(kw in d for kw in ["审计报告", "帮我生成"]): return "llm"
if any(kw in d for kw in ["审计风险", "合规判断"]): return "hybrid"
return "llm"| 模式 | 说明 |
| local | Q0-Q7 并行质量检查、版本锁定、依赖缺口 |
| llm | 审计报告生成、合规判断、增强层选择 |
| hybrid | 本地检查 + LLM 评估 |
阶段0 — 调查(强制入口)
任何审计任务必须先完成调查,产出调查摘要,再进入 Phase-G。
| 维度 | 内容 | 产出 |
| 事实 | 现状/已有文件/配置 | 事实清单 |
| 范围 | 审计边界 | 范围文档 |
| 深度 | 简单/中等/复杂 | 复杂度评级 |
| 依赖 | 外部依赖可用性 | 依赖矩阵 |
| 风险 | 历史失败案例 | 风险清单 |
| 资源 | Token/时间预算 | 资源约束 |
调查方法:workspace 扫描 + rules.md/identity-config.md 读取 + memory_search + sessions_spawn 测试依赖
调查摘要格式:
## 调查摘要
- 事实:{现状}
- 范围:{边界}
- 复杂度:{简单/中等/复杂 + 依据}
- 主要依赖:{依赖项}
- 风险点:{风险}
- 资源约束:{时间/Token/预算}Phase-G — 目标分类(强制,5 步)
| 步骤 | 输出 | 要点 |
| G1 变更分类 | classification | Agent/Skill/系统/跨Agent/进行中/回滚/财务合规 |
| G2 核心问题 | core_problem | 一句话描述 + 目标 |
| G3 成功标准 | success_criteria | 可验证指标 |
| G4 范围确认 | scope | 影响范围 + 边界 + 交叉影响 |
| G5 增强层过滤 | translated_change | A-N 14层自动甄别;confidence≥0.6 升级 |
增强层 A-N 详见 references/enhancement-layers.md。
Layer0 — 质量门槛(前置必做)
| 检查 | 要点 | 参考 |
| Q0 Context 健康 | 3 维度检查 | _knowledge/_components/context-hygiene.md |
| Q1 依赖缺口 | 增强层 A-L 是否遗漏 | — |
| Q2 确认类型 | 是否需要用户确认 | — |
| Q3 复杂度 | 变更是否>5 个 | — |
| Q4 版本锁定 | 版本是否锁定 | frozen_version.json |
| Q5 子代理关联 | workspace 关联正确性 | — |
| Q6 行为红线 | R1-R5 检查 | _knowledge/_components/behavior-checker.md |
| Q7 修复对象就绪 | 单一文件/代码/配置类审计? | 决定是否可输出修复对象 |
Phase-S — 执行框架(概要)
详见 references/execution-framework.md。
| 阶段 | 概要 |
| S1 战略评估 | 8 项准考察 + 风险评级 L1-L4 + 质量属性 QA1-QA5 + 财务合规层(如适用) |
| S2 规划合并 | JSON 规划文档 + 子代理分发 |
| S3 合并审核 | S3a 预检 → S3b Merge Gate → S3c Review Loop(3 次迭代) |
| S4 执行验证 | E5 备份 + 8 项验证 → S4.5 代码实读验证(强制) |
| S5 结果归档 | Artifact 输出 + S5.5 循环限制 → S5.9 进化 → S5.10 经验写入 → S5.11 修复对象输出 |
S3c Review Loop 核心
S3b 发现 P0/P1 fail 时触发,限 3 次迭代(R0/R1/R2)。
S3c.1 收集 fail → S3c.2 派发子代理修复 → [SG1门禁] → S3c.3 验证修复
→ S3c.4 回归 S3b → [SG2门禁] → 通过→S5,未通过→迭代+1SG1 门禁(S3c.2→S3c.3):成功率≥60% + P0 清零 + P1 修复率≥80% SG2 门禁(S3c.4→S5):无新 P0 + 新 P1≤3 + 有 justification
详见 _knowledge/_components/s3c-review-loop-protocol.md。
R1-R5 行为红线(强制)
| 规则 | 简写 | 默认 |
| R1 | 不加不需要的功能 | block |
| R2 | 不过度封装 | warn |
| R3 | 不破坏测试 | block |
| R4 | 不假装测试通过 | block |
| R5 | 先读再改 | warn |
阶段A — 审计(强制)
审计方案完成后、实施前必须执行。详见 references/audit-protocol.md。
| 维度 | 通过标准 |
| 完整性 | Phase-G + S1-S5 齐全,≥3/5 |
| 准确性 | 无虚假描述 |
| 一致性 | 与调查摘要一致 |
| 可执行性 | Task 依赖/步骤清晰 |
| 风险控制 | 风险评级合理 + 降级策略完整 |
Phase-D — 风险评估(可选)
变更步骤>5 时触发:D1 范围 → D2 原因 → D3 缓解 → D4 失效代价
Layer2 — 版本锁定
- 版本文件:
frozen_version.json - 解锁条件:连续 3 次 clean 审计
- 锁定时禁止 P-Sub-P 修改
分级加载索引
| 层级 | 文件 | 加载时机 |
| L1 | SKILL.md(核心+Phase-G+Layer0) | 首次加载 |
| L2 | references/execution-framework.md, references/audit-protocol.md | SKILL.md 后自动 |
| L2 | _knowledge/_components/context-hygiene.md | C 层 (Q0) |
| L2 | _knowledge/_components/behavior-checker.md | B 层 (Q6) |
| L2 | _knowledge/_components/s1-quality-attributes.md | K 层 |
| L2 | _knowledge/_components/financial-compliance.md | L 层(财务合规) |
| L3 | _knowledge/_enhancement/subagent-timeout-recovery.md | 超时/恢复 |
| L3 | _knowledge/_components/s3c-review-loop-protocol.md | S3c 触发 |
| L4 | _knowledge/_components/skill-audit-suite/SKILL.md | skill 审计 |
| L4 | _knowledge/enhancement-registry.md | 专家模式 |
参考文件
| 文件 | 用途 |
references/execution-framework.md | S1-S5 详细执行框架 |
references/audit-protocol.md | 阶段A 审计协议 |
references/enhancement-layers.md | A-N 14 层增强层明细 |
references/self-learning.md | 自学习 + 事实核查 + 优化模块 |
_knowledge/_components/ | 各层组件详见知识库 |
_knowledge/_enhancement/ | 增强层知识注入 |
故障与降级
| 场景 | 处理 |
| 子代理全部超时 | 强制降级主会话(mandatory_override=true) |
| fact_check.py 不可达 | 回退到 exec 验证 |
| SG1 成功率<60% | 主会话接管 + 继续执行 |
| 审计迭代超 5 版 | 强制用户决策 |
| 代码验证 FAIL(S4.5) | 修正结论后重新评分,严禁未验证输出 |
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
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.

