codedb-mcp-fast-code-intelligence

Indexes local repositories with tree-sitter to provide symbol lookup, semantic search, and dependency analysis via MCP tools.

7|2|Updated May 16, 2026
One-click install
npx skills add https://github.com/reason-machines/mcp-skills --skill codedb-mcp-fast-code-intelligence-reason-machines
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: codedb-mcp-fast-code-intelligence
Source: https://github.com/reason-machines/mcp-skills/tree/main/skills/codedb-mcp-fast-code-intelligence
Command: npx skills add https://github.com/reason-machines/mcp-skills --skill codedb-mcp-fast-code-intelligence-reason-machines

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Exploring large codebases with grep or full-file dumps wastes tokens and time. This Skill turns any local repository into a persistent, indexed code intelligence service so agents can answer architecture questions, find symbols, and trace dependencies in milliseconds without reading entire files. ## Core Features & Use Cases - Fast Symbol & Text Search: Tree-sitter and trigram indexing deliver sub-millisecond symbol lookups, outlines, callers, and regex search across thousands of files. - Answer-Oriented Context Tools: codedb_context and codedb_explore return ranked, token-budgeted code excerpts for questions like "how does authentication work?" instead of dumping files. - Dependency & Module Analysis: Analyze import graphs, reverse dependencies, and auto-discovered modules, plus generate a 3D Code Module Atlas and DeepWiki documentation. - Use Case: Before refactoring a payment function, find its definition with codedb_symbol, list all callers with codedb_callers, and check reverse dependencies with codedb_deps to assess impact in seconds. ## Quick Start Ask the agent to index this repository with codedb-mcp and then explain how the authentication flow works using codedb_context.

Frequently Asked Questions about codedb-mcp-fast-code-intelligence

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I set up codedb-mcp for my repository?▼

Build the Rust binary with cargo build --release, create a .codedb-mcp/codedb-mcp.toml config with include/exclude patterns in your project, run codedb-mcp --project-root . index, then register the serve command in your MCP client configuration.

How to search a codebase semantically without reading whole files?▼

Use codedb_search with mode semantic for vector similarity queries, or codedb_context with a natural language question and max_tokens budget to get ranked relevant snippets instead of full file dumps.

Does codedb-mcp support languages other than Rust?▼

Yes, tree-sitter indexing supports multiple languages configured in the [languages] section, including TypeScript, Python, C#, and C++. Enable each language and specify file extensions in codedb-mcp.toml.

Why is my codedb-mcp search slow on the first query?▼

First queries trigger lazy loading of sidecar indexes and the vector embedding model, taking 100ms to 2 seconds. Subsequent warm queries run in sub-millisecond to tens-of-milliseconds range; warm up with a dummy codedb_word or semantic search call.

What to do when codedb-mcp returns stale results?▼

Run an incremental index update with codedb-mcp --project-root . index after code changes. If results remain stale or the cache is corrupted, delete the .bin files and run index --force-rebuild.