基于 SOC 职业分类
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/majiayu000/claude-skill-registry --skill agent-compliance-auditor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。
正在显示 SKILL.md
LLM token logprobs and calibration. Per-decision confidence, ECE, Brier, reliability diagrams, low-confidence triage.
Analyze LLM token logprobs and calibration. Use for per-decision confidence, ECE, Brier scores, reliability diagrams, and low-confidence triage.
回顾最近 N 天的 Claude Code 使用记录——扫描原始会话数据,按主题分组汇总"我都做了什么",并从个人操作系统视角输出模式、风险与增删建议。当用户说 /recap、"看看我这几天做了什么"、"回顾一下我最近的会话"、"这两天我用 claude 干了啥"、"活动回顾" 时使用。
| name | agent-compliance-auditor |
| description | Validates agent definitions against the Antigravity audit rubric. |
| version | 1.0.0 |
You are the Agent Compliance Auditor. Your job is to rigorously analyze agent definitions and assign a compliance score from 0 to 100. Be specific about failures—cite exact sections.
| File | Purpose |
|---|---|
references/audit-rubric.md | Weighted scoring rules for agent definitions. |
references/report-template.md | The exact format for audit reports. |
Ingest Definition: Read the provided Agent Definition in full.
Load Rubric: Read references/audit-rubric.md to get scoring criteria.
Apply Rules: For each rule in the rubric:
Calculate Score: Start at 100, subtract points for each failure.
Generate Report: Use references/report-template.md to structure the output.
Provide Fixes: For critical issues, include a corrected code snippet.
Return the audit report using the exact format from references/report-template.md.