Jeffrey Emanuel Unveils New Multi-Language Code Analysis Tool for AI Agents, Leveraging GPT-5 and Gemini 3

NEW YORK – Jeffrey Emanuel, a prominent figure in AI and open-source development, has announced the release of a new, highly modular code analysis system designed to enhance the capabilities of AI coding agents. The tool, which is part of his "MCP Agent Mail" project, addresses common limitations of traditional linters and type checkers by embracing a higher tolerance for "false positives" when working with advanced large language models (LLMs) like GPT-5 Pro and Gemini 3. Emanuel stated in a recent social media post, "False positives don't matter to [coding agents], except for maybe some wasted tokens checking if it's a legit problem or not. They just need things brought to their attention and they can take it from there." This philosophy underpins the new system, which is built around tools like ast-grep for structural code analysis, moving beyond brittle regex-based methods. The system, initially developed as a Bash script for JavaScript and TypeScript, has rapidly evolved into a comprehensive, modular platform. It now supports a wide array of programming languages, including Python, C++, Rust, Ruby, Swift, Java, and Golang, with easy extensibility for more. Users can install the tool with a single command and run ubs . within any repository to scan projects, with automatic language detection. A key feature is its ability to integrate seamlessly with existing coding agents by providing a ready-made blurb for AGENTS.md or CLAUDE.md files. This integration grants agents "incredible new power" to identify complex and subtle issues that often elude conventional static analysis tools. The MCP Agent Mail project itself acts as a coordination layer, enabling agents to communicate, reserve files, and track progress, akin to an email system for AI. This development highlights a growing trend in AI-assisted software development, where LLMs are increasingly being used not just for code generation but also for sophisticated code analysis and problem-solving, particularly in scenarios requiring contextual understanding and judgment. The tool’s reliance on efficient utilities like ast-grep and ripgrep ensures rapid execution.