Serena
A code intelligence analysis tool based on the Language Server Protocol (LSP). It integrates with AI agents through MCP and supports features such as go-to-definition, find-references and type resolution, improving the accuracy of code comprehension and editing.
Timeline (last 90 days)
The author's core strategy is to have the AI agent cite the output of static analysis tools (ripgrep, ast-grep, Serena, CodeQL) as evidence and to clearly distinguish it from the LLM's own analysis.
Serena provides, through MCP, the language server (LSP) capabilities running behind the IDE, such as go-to-definition and find-references; its advantage is following accurate references and resolving types, but building the index takes time.
In the experiment, Serena mainly played an auxiliary role, used for cross-checking ast-grep and supplementing result verification, and encountered problems such as character encoding conversion and missing type information in external libraries.
All Facts (9)
Serena不面向人类用户，而是面向AI Agent，提供的是程序化的API接口而非图形界面，这与GitHub Copilot、Cursor、Windsurf等工具的设计哲学不同
90%UnverifiedSerena使用Python开发
90%UnverifiedSerena的语义检索能力建立在Language Server Protocol（LSP）之上，LSP最初由微软为VS Code设计
90%UnverifiedSerena的语义编辑使用的底层技术包括tree-sitter（一个增量解析框架，支持数十种编程语言的快速AST构建）和LSP提供的语义分析能力
85%UnverifiedSerena可以被嵌入到LangChain/LangGraph构建的多步骤Agent、AutoGPT类自主代理或企业内部CI/CD自动化流水线中
80%UnverifiedSerena是一个基于MCP（Model Context Protocol）协议的编程工具包，专门为AI编程代理提供语义检索和代码编辑能力
55%UnverifiedThe author's core strategy is to have the AI agent cite the output of static analysis tools (ripgrep, ast-grep, Serena, CodeQL) as evidence and to clearly distinguish it from the LLM's own analysis.
50%UnverifiedSerena provides, through MCP, the language server (LSP) capabilities running behind the IDE, such as go-to-definition and find-references; its advantage is following accurate references and resolving types, but building the index takes time.
50%UnverifiedIn the experiment, Serena mainly played an auxiliary role, used for cross-checking ast-grep and supplementing result verification, and encountered problems such as character encoding conversion and missing type information in external libraries.
50%