An MCP toolkit providing semantic retrieval and editing for AI coding agents.
Give your AI agent symbol-level code retrieval, editing, and refactoring via Model Context Protocol.
Serena is an MCP toolkit that equips AI coding clients like Claude Code, Codex, and IDE assistants with semantic code retrieval and editing. It is built for developers who want their AI agents to operate at the symbol level rather than using primitive text search.
It integrates with any MCP-compatible client to provide high-level abstractions for code exploration, symbol navigation, and refactoring. You can back it with either free Language Server Protocol (LSP) servers supporting over 40 languages or the paid Serena JetBrains Plugin.
It suits developers running terminal-based AI coding agents or IDE assistants who need faster, more reliable code edits on large codebases. It does not suit those not using MCP-compatible clients or projects lacking compatible language servers.
Explore codebases using find symbol and symbol overview instead of reading entire files.
Extends existing AI chat clients and terminal tools via standard MCP server integration.
Choose between open-source language servers covering 40+ languages or the Serena JetBrains Plugin.
Execute symbol renames, moves, inlines, and deletions without fragile text surgery.
The core toolkit relies on open-source language servers, while the Serena JetBrains Plugin is a paid option with a free trial available. The only operational cost is running the server yourself.
880 open remote AI-training roles, updated November 15, 2026; most pay $20–$175 an hour. Zekai lists them free; the application goes to the employer’s platform.
These reviewed AI tools do the same job instantly — from our scored directory.
It supports terminal-based clients like Claude Code and Codex, IDE plugins for VSCode, Cursor, and JetBrains, and desktop chat apps.
It uses either language servers implementing the Language Server Protocol (LSP) or the Serena JetBrains Plugin.
See Zekai first in Google