Skip to main content
MCP server CX Codex

filesift — Claude Code MCP server: code search and setup

filesift indexes a codebase and lets an agent find relevant files by meaning as well as keywords. Use it when a task description is easier to give than a filename.

Install into Claude Code · Platform CLI All methods and platforms ↓
claude mcp add filesift -- uvx filesift
Open on GitHub

What filesift does

The MCP server gives an agent tools to index a project, search it with a natural-language query, find files related to a given file, and check indexing status. Search combines keyword matching with semantic results, so queries can describe code behavior rather than repeat identifiers.

Indexing starts with a structural pass, then builds semantic data in the background. A background daemon serves the indexes, which live in a .filesift directory inside each project. Search runs locally, and the MCP tools communicate over STDIO.

For Claude Code, the listed setup command is claude mcp add filesift -- uvx filesift. The README also describes a filesift-mcp launch command for manual MCP configuration; keep that setup separate from the uvx command.

What it helps with

  • locate authentication handling by its behavior
  • find retry logic for API calls
  • index a project directory for searches
  • trace files related to a given module

Example prompts for the agent

  • Find the code that handles authentication and user sessions.
  • Index this project directory so I can search its code.
  • Find files related to the database connection module.

How to install filesift

Platform CLI

claude mcp add filesift -- uvx filesift

uvx runs the command named after the package; if the README names another one, use uvx --from filesift <command>

Step by step, with commands for each platform: How to add an MCP server to Claude Code, Codex, Cursor, VS Code, Gemini CLI and more

Questions about filesift

How do I install filesift in Claude Code?

Run claude mcp add filesift -- uvx filesift. This adds the MCP server under the name filesift.

What does filesift do?

It indexes a codebase and exposes MCP tools for keyword and semantic search, related-file discovery, and checking index status.

Where does filesift keep project indexes?

It stores each index in a .filesift directory inside the indexed project. Semantic indexing runs in the background after the structural pass.

Similar MCP servers

MCP servers catalog →

Project sponsors

Companies that keep this analytics open
Gold sponsor

Impressions, clicks and code copies — open statistics. We keep the numbers public for the same reason NeBlask itself is public. Become a sponsor