Token导航 LogoToken导航

Agent Skill · 研究检索

claude-permissions-audit(Claude permissions 审核)

用于辅助安全审计、权限检查、凭据风险、认证流程和常见漏洞排查。它适合让 Agent 梳理敏感配置、检查依赖风险、分析鉴权逻辑或生成安全复核清单。使用时不能把工具输出直接当最终结论,涉及密钥、令牌、用户数据或生产系统时,应先确认最小权限、脱敏方式和操作边界。

claude-permissions-audit 是研究检索类 Agent Skill,可用宿主为 Codex、Claude、Cursor、Gemini CLI,通过GitHub 仓库获取,权限为调用外部服务;许可证信息还不明确,第三方扫描有提醒项,按安装量在研究检索类的 73,675 个 Skill 里排第 69,971。

总安装 214周安装 9GitHub Stars 12下载 75核验 2026-05-01

安装说明

本站只整理中文说明,不托管安装包,也不代用户安装。
GitHub
许可证
unknown

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:claude-permissions-audit(Claude permissions 审核)
来源仓库:https://github.com/jackchuka/skills
仓库路径:skills/claude-permissions-audit
安装命令:
npx skills add https://github.com/jackchuka/skills --skill claude-permissions-audit
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 npx skills 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

skills.shnpx skills
npx skills add https://github.com/jackchuka/skills --skill claude-permissions-audit

这个 Skill 是做什么的

研究检索类 Skill 主要解决「找资料、核对事实、把来源整理清楚」的问题,通常会让助手先检索再总结。

用于辅助安全审计、权限检查和凭据风险排查。适合梳理敏感配置、分析鉴权逻辑或生成安全复核清单。不能将输出直接作为最终结论,涉及生产系统时应先确认最小权限和操作边界。通过 npx 从 GitHub 安装,建议检查维护状态和潜在的文件读写行为。

claude-permissions-audit 属于研究检索类 Skill,可作为该场景下的辅助能力补充。

来源仓库内路径:skills/claude-permissions-audit

上手步骤

4 步
  1. 1

    确认自己的宿主工具

    公开资料标注支持 Codex、Claude、Cursor、Gemini CLI;先确认自己用的是其中之一,再按对应工具的 Skills 目录放置。

  2. 2

    拿到安装命令

    可以直接复制的命令是 npx skills add https://github.com/jackchuka/skills --skill claude-permissions-au…;命令里的占位符换成自己的路径或密钥再执行。

  3. 3

    安装后重启宿主

    命令执行完要完全退出宿主工具再打开,让新装的 Skill 被重新扫描到,否则列表里可能看不到。

  4. 4

    按最小权限试用

    它需要调用外部服务,先在测试目录或测试账号里跑,确认没问题再放到正式项目。

常见问题

6 条

claude-permissions-audit 怎么安装?

在终端执行 npx skills add https://github.com/jackchuka/skills --skill claude-permissions-audit。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

claude-permissions-audit 支持哪些 AI 工具?

公开资料标注支持 Codex、Claude、Cursor、Gemini CLI。其他工具只要能读取本地 Skill 目录,一般也能手动放进去,但要自己确认触发方式是否一致。

claude-permissions-audit 需要什么权限?

权限级别是调用外部服务。该 Skill 可能调用第三方服务、云服务或外部模型 API,使用前需要确认账号、额度、数据发送范围和服务条款。

claude-permissions-audit 是免费的吗?

公开资料没有标注明确许可证,收费和商用限制都无法确认。正式使用前请到来源仓库核对授权说明。

claude-permissions-audit 用的人多吗?

公开数据是累计安装 214 次,近一周安装 9 次,下载 75 次,GitHub 12 星。这些数字来自第三方目录统计,只能当作热度参考,不代表质量或安全性。

claude-permissions-audit 安全吗?

第三方扫描有提醒项,说明存在需要人工确认的行为。安装前请打开源码看清它读了什么、联网到哪里。

值得选的地方和要先想清楚的地方

值得选的地方

  • 同时适配 4 种工具,换客户端不用重新找替代品。

先想清楚的地方

  • 许可证没有明确标注,能不能商用、能不能再分发都不确定。
  • 第三方扫描存在提醒项或未通过项,需要人工逐条确认。
  • 第三方目录整理,版本更新和行为细节以来源仓库为准。

适不适合你

更适合

  • 已经在用 Codex,想直接扩展它能力的人。
  • 需要处理研究检索类任务、又不想每次都手写提示词的人。

不太适合

  • 需要明确授权才能用到公司项目里的场景。
  • 需要厂商级技术支持和 SLA 的生产系统。

研究检索类 Skill 对比

6 个同类
Skill总安装GitHub 星权限许可证扫描
find-skills

寻找技能 · Codex、Claude

31,199,81316,486操作浏览器MIT有提醒
self-improving-agent

自我改进 Agent · OpenClaw

16,464,0023,396接触敏感数据MIT-0通过
soultrace

灵魂追踪 · Codex、Claude

7,588,94412需要联网未明确有提醒
self-improving

自我完善 · OpenClaw

6,829,2841,046需要联网MIT-0通过
microsoft-foundry

微软代工厂 · Codex、Claude

6,614,144764调用外部服务未明确有提醒
claude-permissions-audit

Claude permissions 审核 · 本页

21412调用外部服务未明确有提醒

按安装量取同分类前几名;数字来自第三方目录统计,只作热度参考,不代表质量结论。

SKILL.md 原文

Permissions Audit

Review and reorganize Claude Code permission settings so global settings are the source of truth, project-local files stay minimal, and read vs write operations have clear separation.

When to Use

  • Reviewing what tools and commands are pre-allowed
  • Cleaning up accumulated ad-hoc permission entries
  • After adding a new MCP server and deciding which operations to pre-allow
  • Ensuring global vs project-local settings are properly split
  • Periodic hygiene check on permission sprawl

Phase 1: Discover All Settings Files

Scan for every settings file that contributes to the effective permission set.

Locations to check:

  • ~/.claude/settings.json — global settings (source of truth)
  • <project>/.claude/settings.json — project shared settings (committed)
  • <project>/.claude/settings.local.json — project local settings (gitignored)
  • ~/.claude/.claude/settings.local.json — home directory project local
  • Any dotfiles-managed copy (e.g. home/.claude/settings.json in a dotfiles repo)

Read all of them in parallel. If a file doesn't exist, note it and move on.

Phase 2: Audit

For each permission entry across all files, assess:

  1. Redundancy — Is this entry already covered by a broader rule elsewhere?

- Example: Bash(gws calendar events:*) is redundant when Bash(gws:*) exists globally

  1. Misplacement — Is this general-purpose but stuck in a project-local file?

- Example: Bash(python3:*) in project local when it's useful everywhere

  1. Inconsistent granularity — Are similar tools allowed at different specificity levels?

- Example: Global has gh pr list, gh pr view individually, but project local has gh:*

  1. Missing read/write distinction — Are read-only and write operations mixed together?
  2. Stale entries — One-time approvals that accumulated (e.g., a specific git command with hardcoded paths)

Present findings as a table with issues and proposed resolution for each.

Phase 3: Propose Reorganization

Design the new layout following these principles:

Global settings (source of truth)

Organize permissions into labeled sections by grouping related entries together. Do NOT use JSON comments (// or /* */) — settings.json must be valid JSON. Instead, rely on the ordering of entries to convey grouping. Within the allow array, cluster related permissions together in this order:

Section order (group entries in this sequence, no separators needed):

filesystem reads            — cat, tree, grep, wc, file reads
<lang> inspection           — go doc, cargo check, linters (no side effects)
GitHub CLI reads            — gh pr list/view/diff, gh issue list (no mutations)
external tool reads         — calendar, slack CLI, other read CLIs
web reads                   — WebSearch, WebFetch with domain allowlist
MCP <service> reads         — list, get, describe, search operations
build / test / install      — go build, cargo test, pnpm install (local side effects)
skills                      — Skill(superpowers:*), etc.

Project-local settings (minimal overrides)

Only keep entries that are genuinely project-specific:

  • Broader tool access needed for a specific workflow (e.g., Bash(gh:*) for a repo that creates PRs frequently)
  • Project-specific MCP tools not used elsewhere
  • Temporary elevated access during active development

Clarifying questions

Before applying, ask the user about edge cases. Common questions:

  • Scope of CLI tools: Should gh:* (all subcommands) go global, or keep read-only globally with broader access per-project?
  • MCP write operations: Which write operations (send message, create issue) should stay prompted vs pre-allowed?
  • Environment separation: Should dev and prod data sources have the same read access, or should prod be more restrictive?
  • Dotfiles sync: Does the user maintain a dotfiles copy that needs updating?

Adapt questions to what's actually ambiguous — don't ask about things that are obvious from context.

Phase 4: Apply Changes

Once the user confirms the plan:

  1. Write the updated global settings file
  2. Write the trimmed project-local settings file
  3. Update any dotfiles copies to stay in sync
  4. Clear out entries from other local files that were absorbed into global

Phase 5: Summary

Present a before/after comparison:

FileBeforeAfter
GlobalN flat entriesM entries in K sections
Project localX entriesY entries (project-specific only)

List what moved where, and what still requires prompting (write operations not pre-allowed).

平台分布

Codex

38.09%
按下载量换算29

Claude

27.16%
按下载量换算20

Cursor

18.96%
按下载量换算14

Gemini CLI

8.95%
按下载量换算7

安全审计

Gen Agent Trust Hub

可疑

Socket

通过

Snyk

通过

权限和风险

调用外部服务

该 Skill 可能调用第三方服务、云服务或外部模型 API,使用前需要确认账号、额度、数据发送范围和服务条款。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。来源安全扫描存在提醒或未通过项,安装前请自行复核。目前只有一个公开来源,建议同时核对源仓库。

继续浏览同类 Skills