Repo Forensics

What is Repo Forensics?

Repo Forensics is a local security scanner for code you are about to give to an AI agent. The project likens itself to an npm audit for plugins, skills and MCP servers used with AI agents, and it targets the habit of installing third-party extensions for tools such as Claude Code, Codex, OpenClaw and Cursor. You point it at a repository and 29 scanners run in parallel, covering prompt injection, supply chain risk, credential theft, runtime behavior and misconfiguration, with a correlation engine that combines findings into compound threats. Each result carries a confidence score, and the verdict tiers are BLOCK, WARN, INFO and SUPPRESSED, with exit codes for CI gating. Output formats include text, JSON, a compact summary and SARIF 2.1.0. Installed as a plugin, it also monitors installs and updates in the background and blocks known-malicious packages. The README describes it as zero-dependency, with no telemetry and no cloud scanning API. It is an open GitHub project and lists no paid tier.

Use cases

• Vet an agent skill before installing it
• Audit an MCP server for tool poisoning
• Scan a repository for supply chain attacks
• Block known-malicious packages during installs
• Check a project after a supply chain incident
• Gate pull requests in CI with SARIF output

Who it's for

• Developers using coding agents
• Security engineers
• Plugin and skill authors
• DevSecOps teams

Why use Repo Forensics

• Scanning runs on your machine, so the code you audit is not uploaded to a cloud API.
• Zero dependencies and no telemetry keep the tool itself small to trust and review.
• Exit codes and SARIF output let you put the scan into CI and the GitHub Security tab.
• Hooks check installs and updates automatically, so vetting is not a manual step you forget.

Repo Forensics features

• 29 scanners that run in parallel
• Correlation engine with 41 rules for compound threats
• Prompt injection, unicode smuggling and agent config injection checks
• MCP security checks, including tool poisoning and tool shadowing
• Dependency scanning for typosquatting and known CVEs
• Install-hook and lifecycle script analysis
• Git forensics for forged history
• Secrets, SAST, taint-flow and entropy analysis
• Output as text, JSON, summary or SARIF 2.1.0
• Plugin hooks that watch installs, updates and new sessions
• Post-incident scan of node_modules, npm cache and install logs
• Offline mode with signed, self-updating rule packs

Repo Forensics integrations

• Claude Code
• Codex
• OpenClaw
• Cursor
• GitHub Security tab

How to get started with Repo Forensics

1. Install Repo Forensics as an agent plugin or clone the repository from GitHub.
2. Run the run_forensics.sh script against a repository, skill, MCP server or plugin folder.
3. Review the severity-ranked findings and the BLOCK, WARN or INFO verdict.
4. Export the result as JSON or SARIF if you want to use it in CI.
5. Leave the plugin hooks enabled so installs and updates are checked in the background.

Repo Forensics limitations

• Package names, versions and links in scanned repos are looked up unless --offline is used
• Findings come with confidence scores, so ambiguous WARN results need review
• It scans for known and pattern-based threats rather than guaranteeing a repo is safe

Repo Forensics pricing

Repo Forensics is open source, so you can use and self-host it for free.

Prices can change and may vary by country or exclude tax. Check the official pricing page.

Repo Forensics alternatives

The main Repo Forensics alternatives listed on Debuly are SkillSpector, Semgrep, Agent Governance Toolkit, Enola, GitHub MCP Server and skill-optimizer. The table compares pricing and platforms; order is A–Z.

1. 

   Agent Governance Toolkit

   Microsoft's open-source toolkit for enforcing policy, identity, sandboxing and audit records on autonomous AI agents before they reach production.

   Developer platformsAutomation and agents
   Open sourceVisit Agent Governance Toolkit (official site)
2. 

   Enola

   Open-source tool that maps dependencies across repositories and checks each pull request against your architecture rules.

   Developer platformsCoding
   Open sourceVisit Enola (official site)
3. 

   GitHub MCP Server

   GitHub's official MCP server that lets AI assistants read repos, manage issues and pull requests, and watch Actions runs in plain language.

   Developer platformsCodingAutomation and agents
   Open sourceVisit GitHub MCP Server (official site)
4. 

   Semgrep

   Application security platform for static analysis, supply chain and secrets scanning, with AI-assisted triage and fixes for developers and coding agents.

   Developer platformsCoding
   FreemiumVisit Semgrep (official site)
5. 

   skill-optimizer

   Docker workbench and agent skill that runs repeatable, graded evals on agent skills across OpenRouter models.

   Developer platformsCodingAutomation and agents
   Open sourceVisit skill-optimizer (official site)
6. 

   SkillSpector

   NVIDIA's security scanner for AI agent skills that checks for prompt injection, exfiltration and supply-chain risk before install.

   Developer platformsCodingAutomation and agents
   Open sourceVisit SkillSpector (official site)

Repo Forensics FAQ

Sources & verification

Facts on this page come from the official sources below, each with the date we read it. Written by Debuly; the vendor did not review this page.

Image credits

• Logo: none collected yet, so a monogram is shown.
• Cover: none collected yet, so a generated tile is shown.

Spot an error? Report an issue with this listing. We also remove a logo or image on the owner's request.