fuck-u-code
Legacy-Mess Detector – assess the “legacy-mess level” of your code and output a beautiful report
fuck-u-code is an MCP server that assesses code quality across 14 programming languages and exposes a "legacy-mess level" score through AI agent tools. It performs offline static analysis using AST parsing via tree-sitter, scoring code on seven dimensions (complexity, duplication, size, structure, error handling, documentation, naming) and outputting a 0–100 quality score plus per-file "Shit-Gas Index" rankings. The MCP server exposes two tools—analyze and ai-review—allowing AI assistants like Claude Code and Cursor to run quality checks and generate AI-powered reviews directly within their workflows. It also includes an optional agent skill that helps AI agents interpret the analysis output and produce structured feedback.
Key Features
Use Cases
- 01Enabling Claude Code or Cursor to assess legacy codebases and surface the most problematic files
- 02Generating automated code quality reports during PR reviews via AI assistant workflows
- 03Identifying technical debt hotspots across multi-language monorepos
- 04Running AI-driven refactoring suggestions on the worst-scoring files in a project
- 05Exporting quality metrics as HTML dashboards or JSON for CI/CD pipeline integration
- 06Teaching AI agents to interpret and act on code quality scores using the bundled skill
Related MCP Servers
View moremarkitdown
Python tool for converting files and office documents to Markdown.
firecrawl
The context API to search, scrape, and interact with the web at scale. 🔥
prompts.chat
f.k.a. Awesome ChatGPT Prompts. Share, discover, and collect prompts from the community. Free and open source — self-host for your organization with complete privacy.
langflow
Langflow is a powerful tool for building and deploying AI-powered agents and workflows.
fuck-u-code — FAQ
What is the fuck-u-code MCP server?+
It's an MCP server that performs static code quality analysis across 14 languages and exposes two tools (analyze and ai-review) to AI assistants like Claude Code and Cursor. The analysis runs fully offline; AI review requires an external LLM API or local Ollama.
How do I install and connect fuck-u-code to Claude Code or Cursor?+
Install globally with 'npm install -g eff-u-code', then run 'fuck-u-code mcp-install' and choose your client (claude or cursor), or manually add the MCP server entry to ~/.claude.json or .cursor/mcp.json with command 'fuck-u-code-mcp'.
Which AI clients work with this MCP server?+
Claude Code, Cursor, and any MCP-compatible client (e.g., Windsurf). The README documents auto-configuration for Claude Code and Cursor.
Do I need API keys to use fuck-u-code?+
The analyze tool requires no API keys and runs offline. The ai-review tool requires an API key for OpenAI-compatible, Anthropic, DeepSeek, or Gemini providers, or a local Ollama installation.
Is fuck-u-code free and open source?+
Yes, it's MIT-licensed and available on npm as 'eff-u-code'. The MCP server and CLI are both included in the same package.
How do I install the agent skill component?+
Use 'npx degit Done-0/fuck-u-code/skills/fuck-u-code-analysis ~/.claude/skills/fuck-u-code-analysis' (adjust path for opencode or other agents). This requires the global npm package 'eff-u-code' to be installed first.
How do I install fuck-u-code?+
Open the source repository on GitHub and follow its README. fuck-u-code is a mcp server — MCP Agents Market links you directly to the official repo.
Is fuck-u-code free?+
fuck-u-code is an open-source project hosted on GitHub. Check the repository for its license and any usage requirements.