Amazon Q Developer
Generative AI coding assistant for building, operating, and transforming software
Open-Source Projects
Discover top open-source software, updated regularly with real-world adoption signals.
Discover top open-source software, updated regularly with real-world adoption signals.
Turn any LLM into a code‑aware development agent
Serena equips LLMs with IDE‑like semantic code retrieval and editing, supporting dozens of languages and integrating via MCP with many clients, all free and open‑source.
Serena is designed for developers and teams that rely on LLM‑powered coding assistants to work on medium‑to‑large codebases. It addresses the inefficiency of having agents read whole files or perform brittle grep‑style searches, delivering precise, symbol‑level operations.
By leveraging language‑server‑protocol (LSP) back‑ends, Serena provides tools such as find_symbol, find_referencing_symbols, and insert_after_symbol. These enable IDE‑like navigation, retrieval, and editing across more than twenty programming languages. The toolkit is LLM‑agnostic and integrates through a Model Context Protocol (MCP) server, making it compatible with Claude Code, Claude Desktop, VSCode, Cursor, terminal‑based clients, and custom agent frameworks.
Run the MCP server locally, via Docker, Nix, or as a command‑line utility, then connect your preferred LLM client. Because the toolset is decoupled from any specific framework, you can embed it in existing pipelines or extend it with new language adapters without altering core logic.
When teams consider Serena, these hosted platforms usually appear on the same shortlist.
Looking for a hosted option? These are the services engineering teams benchmark against before choosing open source.
Token‑efficient refactoring in Claude Code
Agent locates and updates target symbols across a 200k‑line repo, cutting token consumption by ~40%.
Cross‑language bug fix via VSCode extension
Developer invokes Serena tools to find referencing symbols in Python and TypeScript, applying fix in seconds.
Automated feature addition in CI pipeline
CI job runs Serena MCP server to insert new logging after a specific function, ensuring consistent code changes without manual edits.
Custom agent integration for internal tooling
Team embeds Serena's API into their proprietary agent framework, gaining IDE‑like capabilities without rewriting existing logic.
No, Serena is LLM‑agnostic; it provides tools that any model can call via MCP or OpenAPI.
Serena leverages LSP and includes ready support for Python, TypeScript/JavaScript, PHP, Go, R, Rust, C/C++, Zig, C#, Ruby, Swift, Kotlin, Java, Clojure, Dart, Bash, Lua, Nix, Elixir, Erlang, and more.
Run the provided MCP server locally, via Docker, Nix, or as a command‑line tool; then connect your LLM client.
Yes, it is released under an open‑source license and can be used commercially at no cost.
Java language server startup is slower and may have platform‑specific issues, especially on macOS and Linux, which are being addressed.
Project at a glance
ActiveLast synced 4 days ago
