Serena review
Serena is an MCP toolkit that gives coding agents semantic, symbol-level code retrieval, editing, refactoring and debugging tools, similar to an IDE.
TL;DR
Serena is an MCP server that gives AI coding agents IDE-like tools that work on symbols rather than line numbers or text patterns. It is for developers using MCP clients such as Claude Code, Codex or Cursor. Serena runs on open-source language servers by default, or on a paid JetBrains plugin with extra refactoring and debugging tools.
- 31kGitHub stars
- +307new stars, week of Oct 25
- 35kdownloads a week, −4% vs 4-week avg
- #35in MCP and agent tools this week
Serena at a glance
- Category
- MCP and agent tools
- Maintained by
- Oraios AI (organization, @oraios on GitHub)
- Language
- Python, JavaScript, HTML
- Repository
- github.com/oraios/serena
- Website
- oraios.github.io/serena
- PyPI package
- serena-agent 2.1.0
- Latest release
- v2.1.0 on GitHub, Oct 19, 2026
- First published
- Mar 23, 2025
What Serena does
- Serena finds symbols, file outlines, referencing symbols, declarations and implementations so agents can explore code without reading entire files.
- Serena edits code at the symbol level: replace a symbol body, insert before or after a symbol, and safe delete.
- Serena renames symbols. With the JetBrains plugin, it also renames files and directories, moves code, inlines, and removes unused code.
- Serena supports over 40 programming languages through its language server backend.
- Serena offers interactive debugging with breakpoints, variable inspection and expression evaluation, but only with the JetBrains plugin.
- Serena includes a memory system for sharing knowledge across sessions, users and projects. You can disable it.
- Serena is configured in YAML at global, CLI, per-project, client-context and mode levels.
How it works
Serena is an MCP server that adds tools to your existing AI client, while an LLM orchestrates the work. You can give the client a launch command, or start Serena in HTTP mode and give the client its URL. Code analysis comes from LSP language servers by default, or from your JetBrains IDE through the Serena JetBrains Plugin. Serena v2 adds a REPL-based agent interface.
How to install and start using Serena
With uv
uv tool install -p 3.13 serena-agentRequirements
- uv must be installed to install and manage Serena.
- The PyPI package requires Python >=3.11 and <3.15.
- An MCP-capable client with an LLM is required to orchestrate Serena's tools.
- Some languages need extra dependencies when you use the language server backend.
First steps
- Install uv, which is the only required prerequisite, then install Serena with uv.
- Run serena init to initialize Serena and verify your setup. Add -b JetBrains to use the JetBrains backend.
- Configure a launch command for Serena in your MCP client, such as Claude Code, Codex or Claude Desktop.
- Read the user guide on Serena's project-based workflow and configuration.
Who Serena is for
- Developers using coding agents on large or complex codebases who need cross-file navigation and refactoring.
- Users of Claude Code, Codex, Copilot, Cursor, Gemini-CLI or JetBrains AI assistants who want symbol-aware tools.
- Multi-language projects and monorepos, across more than 40 supported languages.
- JetBrains IDE users who want agents to use the IDE's refactoring and debugging features.
Limitations to know
- Serena's move, inline, type hierarchy, dependency search and debugging tools require the paid JetBrains plugin.
- The Serena JetBrains Plugin does not support Rider or CLion.
- With language servers, find declaration generally fails for external dependencies, and find implementations works only for some languages.
- The README warns not to install Serena from MCP or plugin marketplaces because their install commands are outdated.
Is Serena growing? Usage data
In the week of Oct 25–31, 2026, Serena ranked #244 on the Devtoolwatch Board and #35 in MCP and agent tools, with a momentum score of 38 out of 100. Its main package had 35k downloads that week, −4% against its own 4-week average.
Show the weekly numbers
| Week of | New stars |
|---|---|
| Aug 9, 2026 | 335 |
| Aug 16, 2026 | 329 |
| Aug 23, 2026 | 266 |
| Aug 30, 2026 | 273 |
| Sep 6, 2026 | 385 |
| Sep 13, 2026 | 389 |
| Sep 20, 2026 | 222 |
| Sep 27, 2026 | 162 |
| Oct 4, 2026 | 232 |
| Oct 11, 2026 | 213 |
| Oct 18, 2026 | 356 |
| Oct 25, 2026 | 307 |
The Board ranks tools by four signals each week: new stars, new stars for the repo's size, added downloads and download growth beyond the market. How the ranking works.
Serena alternatives
Other MCP servers and agent tools on the Board in the week of Oct 25–31, 2026, by momentum:
| Rank | Tool | Stars | Downloads/wk |
|---|---|---|---|
| #1 | REAREA is an MCP server and CLI that lets coding agents reverse engineer native binaries, JavaScript and Electron apps, .NET assemblies,… | 415.5k | 760.6k |
| #2 | ha-mcpha-mcp is an unofficial Model Context Protocol server that lets AI assistants control and manage Home Assistant using natural language. | 5.4k | 118.1k |
| #3 | TregTreg is a registry and proxy that gives AI agents one token for a catalog of paid third-party API endpoints and team-shared tools. | 6.9k | 6.7k |
| #4 | Chrome DevTools MCPChrome DevTools MCP is an MCP server and CLI that lets coding agents control and inspect a live Chrome browser using Chrome DevTools. | 54.6k | 2.7M |
| #5 | codebase-memory-mcpHigh-performance code intelligence MCP server. Indexes codebases into a persistent knowledge graph — average repo in milliseconds. 162… | 47.9k | – |
| #6 | open-computer-use👾 Open Computer Use – Open-Source Alternative to Codex Computer Use | 2.8k | – |
Recent Serena releases
Serena FAQ
Is Serena free and open source?
Serena's code is on GitHub (oraios/serena), but the repository doesn't declare a standard open-source license. Check its license file before using it commercially.
How do I install Serena?
Run uv tool install -p 3.13 serena-agent. The serena-agent package needs Python 3.11 or newer, below 3.15. The install section above has the other options from the project's README.
Is Serena still maintained?
Yes. The oraios/serena repository was last updated on Oct 20, 2026, and the latest release, v2.1.0 on GitHub, came out on Oct 19, 2026.
How popular is Serena?
Serena has 31k GitHub stars as of Nov 1, 2026. It gained 307 new stars in the week of Oct 25–31, 2026, against a 12-week average of 289 a week. Its main package had 35k downloads that week, −4% against its own 4-week average. That made it #35 among MCP servers and agent tools on the Devtoolwatch Board.
What are the best alternatives to Serena?
Other MCP servers and agent tools on the Devtoolwatch Board in the week of Oct 25–31, 2026, by momentum: REA (#1), ha-mcp (#2), Treg (#3), Chrome DevTools MCP (#4), codebase-memory-mcp (#5). The alternatives table above has their numbers.
How do I use Serena with Claude Code?
Serena connects to Claude Code as an MCP server. Install Serena with uv, run serena init, then configure a launch command for Serena in Claude Code. The documentation's client page has specific instructions. Inside harnesses like Claude Code, Serena's basic file and shell tools are typically disabled by default.
How do I use Serena MCP?
Serena MCP works in one of two ways. Either your client starts the Serena server through a launch command, or you start Serena in HTTP mode and give the client its URL. An LLM in the client then calls Serena's tools.
Does Serena require the JetBrains plugin?
Serena does not require the JetBrains plugin. It uses free, open-source language servers by default. The paid JetBrains plugin, which has a free trial, adds move, inline, type hierarchy and debugging tools.
Which clients work with Serena?
Serena works with MCP clients. These include Claude Code, Codex, OpenCode, Gemini-CLI, VSCode, Cursor and JetBrains assistants, as well as Claude Desktop, Codex App and OpenWebUI.
Sources
Serena README badge
Maintainers can show the tool's weekly Board rank in their README:
[](https://index.devtoolwatch.com/tools/serena/)How this page is made: Devtoolwatch writes each review from the project's README, releases, package metadata and website, with an AI model, then checks every claim against the source it cites and drops what the sources don't support. The numbers come from Devtoolwatch's daily GitHub snapshots, GitHub's stargazer history, and the npm and PyPI download APIs. Spotted an error? Email [email protected].