Function Signature Lookup MCP Server

The Function Signature Lookup MCP Server fills a crucial gap for developers building AI‑powered coding assistants: the ability to query, in real time, precise function prototypes from a project's source tree. When an assistant needs to suggest or validate API calls, it can call on this server and receive a detailed signature—parameter types, return type, and documentation tags—without having to parse the code itself. This eliminates the need for custom language parsers or maintaining separate documentation services, letting developers focus on higher‑level logic.

At its core, the server uses a lightweight SQLite index populated by Universal Ctags. During indexing, ctags scans the project’s source files (honoring to skip irrelevant paths) and extracts tags for functions, classes, and other callable entities across all languages supported by ctags. The resulting JSON‑formatted tag data is stored in the database, enabling fast lookups even for large codebases. Developers can trigger re‑indexing on demand with , or drop pre‑generated ctags files into the directory to refresh the index without restarting.

Key capabilities include:

• Cross‑language support: Works seamlessly with Python, C, Go, Rust, JavaScript, Java, and more.
• Fine‑grained queries: Retrieve signatures by function name, list all APIs in a file, or enumerate files belonging to an API.
• Integration hooks: The MCP functions (, , etc.) expose a clean, typed interface that Claude or other assistants can call directly.
• Extensibility: The server accepts new ctags files on the fly, making it easy to keep the index current as a codebase evolves.

Typical use cases include:

• Autocomplete and documentation lookup for IDE‑style assistants that need to surface accurate type hints.
• Code review bots that verify function signatures against specifications or detect deprecated APIs.
• API migration tools that map old function names to new ones across a multi‑language repository.
• Automated test generation where the assistant must understand function contracts to create meaningful stubs.

By decoupling signature extraction from the AI client, this MCP server provides a lightweight, language‑agnostic service that plugs directly into existing AI workflows. Its reliance on universally supported tools (Python, SQLite, Universal Ctags) ensures portability and ease of deployment, while its rich query surface empowers developers to build smarter, context‑aware assistants without reinventing parsing logic.