Fossil MCP – Dead code detection for vibe-coded projects (github.com)

Fossil MCP has launched a static analysis toolkit specifically designed for projects where AI-driven "vibe coding" is prevalent. As developers increasingly rely on AI tools like GitHub Copilot to generate code, the proliferation of dead code, duplicated logic, and leftover scaffolding becomes a significant challenge, often going unnoticed. Fossil MCP addresses these issues by detecting unreachable functions, duplicated utilities, and residual temporary markers left behind by AI sessions across 15 programming languages. Unlike traditional linters, Fossil MCP utilizes a call graph to understand how functions are interconnected, allowing for granular traversal and analysis of the codebase. This tool is significant for the AI/ML community as it enhances code quality in scenarios where developers may overlook critical aspects of the AI-generated code. With features like zero configuration setup, automatic framework detection, and efficient artifact identification, Fossil MCP not only helps maintain a clean code environment but also reduces developer time spent debugging and refining AI output. Built in Rust for speed and memory safety, it serves as both a command-line interface tool and an MCP server, enabling AI agents to check their own outputs, thereby streamlining the coding process while minimizing costs associated with repeated AI inference.