proofbench
llms.txt1 starsChanged 12 days agoMITView on GitHub
Open-source Agent Skills and MCP servers for AI coding agents (Claude Code, GitHub Copilot, Codex, Cursor, Gemini CLI): bug analysis, regression tracing with git bisect, proven bug fixes, feature development, code cleanup, security audits, Android APK reverse engineering and ticket reading. Skills trigger automatically from plain sentences; scores are computed by MCP servers from recorded evidence.
In this repository9 other files it gives its agents.
- AGENTS.mdAGENTS.md
- analyze-apkSkill
- build-featureSkill
- clean-upSkill
- fix-bugSkill
- read-ticketSkill
- regression-traceSkill
- security-auditSkill
- using-proofbenchSkill
The file
Files
# Proofbench > Open-source Agent Skills and MCP servers for AI coding agents (Claude Code, GitHub Copilot, Codex, Cursor, Gemini CLI): bug analysis, regression tracing with git bisect, proven bug fixes, feature development, code cleanup, security audits, Android APK reverse engineering and ticket reading. Skills trigger automatically from plain sentences; scores are computed by MCP servers from recorded evidence. ## Docs - [README](README.md): what it is, install for each assistant, FAQ, safety model - [Skills](skills/README.md): the seven skills and the router - [APK analysis guide](docs/apk-analysis.md): the 14 apk-inspector tools, bytecode and caller-check analysis, optional jadx bridge, limits - [apk-inspector](mcp/apk-inspector/README.md): MCP server for static APK analysis - [regression-tracer](mcp/regression-tracer/README.md): MCP server for regression investigations - [workflow-gate](mcp/workflow-gate/README.md): MCP server for proof-backed workflows - [Vision](docs/VISION.md): promise, target, risks ## Optional - [Design specs](docs/design/specs/2026-11-02-regression-tracer-design.md) - [Contributing](CONTRIBUTING.md)
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.