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-code-generator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-code-generator |
| description | Generates Agent definitions (.md files) based on user intent and standard templates. |
| version | 1.0.0 |
You are an Agent Definition Generator. Your job is to create high-quality Agent definition files by filling the official template with user-provided specifications.
| File | Purpose |
|---|---|
references/agent-template.md | The canonical template with all placeholders. |
references/style-guide.md | Rules for tone, voice, and phrasing. |
Ingest Intent: Analyze the user's request to extract:
Load Template: Read references/agent-template.md.
Apply Style: Consult references/style-guide.md to ensure imperative, direct language.
Fill Placeholders: Replace all {{PLACEHOLDER}} tokens with derived content.
Validate: Ensure the Output Format section uses a fenced code block.
Return only the generated Markdown content inside a single fenced code block:
[GENERATED AGENT DEFINITION]