claude-code-repo-audit
llms.txt2 starsChanged 4 days agoMITView on GitHub
Repo Audit is a Claude Code skill and plugin that audits an untrusted repository without running it. It acquires the code into quarantine, maps every place code runs automatically (install scripts, IDE tasks, git config, dev containers, tool configs, AI-agent settings), triages dependencies against registry metadata, hunts remote-code loaders, obfuscation, hidden Unicode and blockchain-hosted payloads, checks credential and wallet access, reads git history, and can check the host for persistence. It returns CLEAN, SUSPICIOUS or MALICIOUS with path:line evidence. Install:…
In this repositoryOne other file it gives its agents.
- repo-auditSkill
The file
Files
# Repo Audit > Repo Audit is a Claude Code skill and plugin that audits an untrusted repository without running it. It acquires the code into quarantine, maps every place code runs automatically (install scripts, IDE tasks, git config, dev containers, tool configs, AI-agent settings), triages dependencies against registry metadata, hunts remote-code loaders, obfuscation, hidden Unicode and blockchain-hosted payloads, checks credential and wallet access, reads git history, and can check the host for persistence. It returns CLEAN, SUSPICIOUS or MALICIOUS with path:line evidence. Install: `/plugin install repo-audit --marketplace albkhachikyan/claude-code-repo-audit`. Invoke: `/repo-audit:repo-audit <git URL | path | zip>` (plugin) or `/repo-audit <…>` (manual install). Quick mode: `--quick`. License: MIT. Author: Albert Khachikyan. ## Docs - [README](https://github.com/albkhachikyan/claude-code-repo-audit/blob/main/README.md): overview, install, usage, checks, comparison, FAQ - [SKILL.md](https://github.com/albkhachikyan/claude-code-repo-audit/blob/main/SKILL.md): the phases and hard rules the skill follows - [Execution surfaces](https://github.com/albkhachikyan/claude-code-repo-audit/blob/main/references/execution-surfaces.md): every place code runs without a choice, per ecosystem - [Malicious patterns](https://github.com/albkhachikyan/claude-code-repo-audit/blob/main/references/malicious-patterns.md): code patterns and de-fanged indicators - [Dependency triage](https://github.com/albkhachikyan/claude-code-repo-audit/blob/main/references/dependency-triage.md): registry checks and tarball triage - [Social engineering](https://github.com/albkhachikyan/claude-code-repo-audit/blob/main/references/social-engineering.md): lure red flags for fake interviews - [Host compromise](https://github.com/albkhachikyan/claude-code-repo-audit/blob/main/references/host-compromise.md): host persistence checks per OS - [Incident response](https://github.com/albkhachikyan/claude-code-repo-audit/blob/main/references/incident-response.md): steps after MALICIOUS or after something ran - [Case study, 2026-11](https://github.com/albkhachikyan/claude-code-repo-audit/blob/main/references/case-study-2026-11.md): the worked real case behind the skill - [Audit report template](https://github.com/albkhachikyan/claude-code-repo-audit/blob/main/templates/audit-report.md): report format for phase 8 ## Optional - [Task Splitter](https://github.com/albkhachikyan/claude-code-task-splitter): companion skill that routes subtasks to the cheapest capable model - [Next.js Frontend Foundation](https://github.com/albkhachikyan/nextjs-frontend-foundation): companion skill that scaffolds strict Next.js 16 frontends
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.