by xyiqq
Quality gate for Agent Skills: lint, security audit, and Claude/Cursor/Codex/OpenCode compatibility.
# Add to your Claude Code skills
git clone https://github.com/xyiqq/skilldoctorLast scanned: 8/15/2026
{
"issues": [
{
"file": "README.md",
"line": 134,
"type": "remote-install",
"message": "Install command (remote install script piped to a shell — review the source before running): \"curl | sh\"",
"severity": "low"
},
{
"file": "test/fixtures/audit-injection/SKILL.md",
"line": 8,
"type": "prompt-injection",
"message": "Possible instruction-override phrase: \"Ignore previous instructions\"",
"severity": "medium"
},
{
"file": "test/fixtures/audit-injection/SKILL.md",
"line": 8,
"type": "prompt-injection",
"message": "Possible concealment directive: \"do not tell the user\"",
"severity": "medium"
},
{
"file": "test/fixtures/audit-pipe-shell/SKILL.md",
"line": 8,
"type": "remote-install",
"message": "Install command (remote install script piped to a shell — review the source before running): \"curl https://example.com/install.sh | bash\"",
"severity": "medium"
}
],
"status": "PASSED",
"scannedAt": "2026-08-15T04:30:02.264Z",
"npmAuditRan": true,
"pipAuditRan": true,
"promptInjectionRan": true
}See how skilldoctor compares with popular alternatives.
skilldoctor is an open-source ai agents skill for AI coding assistants such as Claude Code, Codex CLI, and ChatGPT, built by xyiqq. Quality gate for Agent Skills: lint, security audit, and Claude/Cursor/Codex/OpenCode compatibility. It has 434 GitHub stars.
Yes. skilldoctor passed SkillsLLM's automated security scan — a dependency vulnerability audit plus prompt-injection heuristics — with no high-severity issues. You can read the full report in the Security Report section on this page.
Clone the repository with "git clone https://github.com/xyiqq/skilldoctor" and add it to your Claude Code skills directory (see the Installation section above).
skilldoctor is primarily written in TypeScript. It is open-source under xyiqq on GitHub, so you can review or fork the full source.
Yes. SkillsLLM lists many other AI Agents skills you can browse and compare side by side. Open the AI Agents category from the badge at the top of this page, or use the Related Skills and comparison links further down to weigh skilldoctor against similar tools.
No comments yet. Be the first to share your thoughts!
⚠️ Third-Party Software Notice
This skill is third-party open-source software developed and hosted independently on GitHub. SkillsLLM is an informational directory and does not control or maintain the underlying repository.
Any security checks, ratings, or warnings displayed by SkillsLLM are automated and limited in scope. They do not constitute a security certification or guarantee that the software is safe, error-free, or free from malicious code, vulnerabilities, compromised dependencies, or prompt-injection risks.
Review the source code, permissions, dependencies, and configuration before installing or running any third-party skill. Use is at your own risk. To the maximum extent permitted by applicable law, SkillsLLM is not liable for losses arising from third-party software.
Quality gate for Agent Skills. Lint the spec, audit unsafe instructions, and check whether a SKILL.md actually works on Claude Code, Cursor, Codex, OpenCode, Gemini CLI, and GitHub Copilot.
Agent Skills 的质量门禁:校验官方规范、审计危险指令,并检查同一个 SKILL.md 在 Claude Code、Cursor、Codex、OpenCode、Gemini CLI、GitHub Copilot 上能不能移植。
Vercel npx skills installs skills. skilldoctor decides whether you should keep them.
Vercel 的 npx skills 负责安装。skilldoctor 负责判断该不该留。
npx --yes github:xyiqq/skilldoctor --help
npx --yes github:xyiqq/skilldoctor lint .
npx --yes github:xyiqq/skilldoctor audit .
npx --yes github:xyiqq/skilldoctor compat .
npx --yes github:xyiqq/skilldoctor ci .
From a clone / 从源码:
git clone https://github.com/xyiqq/skilldoctor.git
cd skilldoctor
npm install
npm run build
node dist/cli.js --help
Requires Node.js 18.18 or newer.
npx --yes github:xyiqq/skilldoctor lint ./my-skill
npx --yes github:xyiqq/skilldoctor audit ./my-skill
npx --yes github:xyiqq/skilldoctor compat ./my-skill
npx --yes github:xyiqq/skilldoctor ci ./my-skill
npx --yes github:xyiqq/skilldoctor scan
npx --yes github:xyiqq/skilldoctor init pdf-processing
npx --yes github:xyiqq/skilldoctor rules
npx --yes github:xyiqq/skilldoctor explain lint/name-invalid
npx --yes github:xyiqq/skilldoctor fix ./my-skill
npx --yes github:xyiqq/skilldoctor fix ./my-skill --dry-run
npx --yes github:xyiqq/skilldoctor score ./my-skill
scan is also available as skilldoctor doctor.
path can be one skill directory or a repository root. skilldoctor walks the tree and finds every SKILL.md.
path 可以是单个 Skill 目录,也可以是仓库根目录。skilldoctor 会向下查找全部 SKILL.md。
| Flag | Values | Default |
|---|---|---|
--format |
human, json, sarif, markdown |
human |
--fail-on |
error, warning, never, or score:<n> for score |
error |
--ignore |
glob or path prefix, repeatable | none |
--suppress |
rule id or lint/* style prefix, repeatable |
none |
--output |
file path to write the report | none |
--quiet |
off | |
--dry-run |
with fix only |
off |
skilldoctor ci . --format json --fail-on warning
skilldoctor ci . --ignore examples --format markdown
skilldoctor ci . --suppress lint/description-vague --suppress compat/*
skilldoctor ci . --format markdown --output skilldoctor-report.md
skilldoctor score . --fail-on score:80
skilldoctor scan --ignore keep-out
Optional config files in the repo root / 仓库根目录可选配置:
{
"failOn": "error",
"ignore": ["vendor/skills"],
"suppress": ["lint/description-vague", "compat/*"]
}
.skilldoctorignore uses one path prefix per line. # comments are skipped.
.skilldoctorignore 每行一个路径前缀,# 开头是注释。
suppress hides matching findings from reports, exit codes, scores, and GitHub annotations. Use it for gradual adoption, not to silence security errors forever.
suppress 会从报告、退出码、评分和 GitHub annotation 里隐藏匹配规则。适合渐进接入,不要长期用来掩盖安全问题。
skilldoctor ci 1 skill(s)
✔ release-notes examples/release-notes
1 skills 0 errors 0 warnings 0 info
A failing audit looks like this / 失败的审计类似:
✖ audit-injection test/fixtures/audit-injection
error audit/prompt-injection SKILL.md:8 instructions try to override system or hidden-user policy
name: skilldoctor
on:
pull_request:
push:
branches: [main]
jobs:
skills:
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v4
- uses: actions/setup-node@v4
with:
node-version: 22
- uses: xyiqq/skilldoctor@v0.2.4
with:
path: .
fail-on: error
suppress: lint/description-vague
output: skilldoctor-report.md
Or run the CLI from this repository / 或者直接跑本仓库 CLI:
- run: npx --yes github:xyiqq/skilldoctor ci .
name, description, directory match, YAML, metadata types, 500-line budget, broken references/ / scripts/ / assets/ linksBash(*), curl | sh, dump hosts, self-modifying skillsRule IDs are stable: docs/RULES.en.md / docs/RULES.zh-CN.md
The installer slot is taken. This tool is the missing CI gate: authors add one command, reviewers see annotations, users run audit before a skill touches ~/.codex/skills.
安装器赛道已经有人占了。这个工具补的是 CI 门禁:作者加一条命令,审查者能在 PR 里看到 annotation,用户在 Skill 进 ~/.codex/skills 之前先跑 audit。
npm install
npm test
npm run build
node dist/cli.js lint examples
See CONTRIBUTING.md.
MIT