CodeGraph

What is CodeGraph?

CodeGraph builds a pre-indexed knowledge graph of the symbols, call edges and dependencies in your codebase and serves it to AI coding agents through an MCP server. Instead of grepping and opening files one by one, the agent asks one question and gets the relevant source, the call paths between symbols and the blast radius of a change. Parsing runs in a native Rust kernel for 20 languages, the index lives in a local SQLite database, and a file watcher keeps it current as code changes. Installation is a single curl or PowerShell command, or npm, followed by codegraph install to wire up agents and codegraph init in each project. Across seven benchmark repositories, the README reports average savings of 62% in tokens and 44% in cost, and it notes a larger footprint in the context window. It is MIT licensed, with a hosted platform on a waitlist.

Use cases

• Speed up code questions in AI agents
• Reduce tool calls and tokens in coding agents
• Trace callers and callees of a symbol
• Check the impact radius of a change
• Search a codebase by name with full-text search
• Index mixed iOS and React Native projects

Who it's for

• Developers using AI coding agents
• Open-source contributors
• Engineering teams
• Agent builders

Why use CodeGraph

• Agents get relevant code and call paths in one request, which cuts the discovery loop of grep and file reads.
• Everything runs on your machine with a SQLite database, so no code is sent to an outside service.
• Auto-sync keeps the index fresh while an agent or you edit files.
• One installer configures many agents, so a team can share the same setup.
• It is free under the MIT license.

CodeGraph features

• Pre-built graph of symbols, calls and dependencies
• MCP server for coding agents
• Native Rust parsing kernel for 20 languages
• Full-text search powered by FTS5
• Impact analysis of callers, callees and affected code
• File watcher with automatic debounced sync
• Framework-aware route detection across 17 frameworks
• Cross-language flows for Swift, Objective-C and React Native
• CLI and library API
• Fully local with no API keys

CodeGraph integrations

• Claude Code
• Cursor
• Codex CLI
• OpenCode
• Hermes Agent
• Gemini CLI
• Antigravity
• Kiro
• GitHub Copilot

How to get started with CodeGraph

1. Install the CLI with the curl or PowerShell one-liner, or with npm i -g @colbymchenry/codegraph.
2. Run codegraph install in a new terminal to wire the MCP server into the agents it detects.
3. In each project, run codegraph init to create the local .codegraph directory and build the graph.
4. Your agent queries the graph through MCP, and the watcher updates the index as files change.

CodeGraph limitations

• Benchmarks come from the project's own tests on seven repositories
• Per the README, its answers leave about 80% more retrieval context in the window at session end
• Each project must be initialized with codegraph init
• The hosted CodeGraph platform is only available through a waitlist

CodeGraph pricing

CodeGraph is open source, so you can use and self-host it for free.

Prices can change and may vary by country or exclude tax. Check the official pricing page.

CodeGraph alternatives

The main CodeGraph alternatives listed on Debuly are code-review-graph, Repomix, Greptile, Bodega One Code, ContextStream and Graft (now Trail-code). The table compares pricing and platforms; order is A–Z.

1. 

   Bodega One Code

   Free local-first AI coding agent and full IDE that runs parallel background agents on your machine with any local or cloud model you bring.

   CodingDeveloper platformsAutomation and agents
   FreeVisit Bodega One Code (official site)
2. 

   code-review-graph

   Open-source local code graph and MCP server that gives AI coding tools only the context a diff touches, instead of the whole repo.

   CodingDeveloper platforms
   Open sourceVisit code-review-graph (official site)
3. 

   ContextStream

   Shared project memory, code search and agent coordination for AI coding tools, connected through MCP and a browser workspace.

   CodingDeveloper platformsAutomation and agents
   FreemiumVisit ContextStream (official site)
4. 

   Graft (now Trail-code)

   Open source context layer that maps a codebase for coding agents and stores team learnings as Markdown files in git.

   CodingDeveloper platformsAutomation and agents
   Open sourceVisit Graft (now Trail-code) (official site)
5. 

   Greptile

   AI pull request reviewer that maps your whole repository, so each review weighs the change against code outside the diff.

   CodingDeveloper platforms
   FreemiumVisit Greptile (official site)
6. 

   Repomix

   Open-source tool that packs a whole code repository into one AI-friendly file with token counts and secret checks.

   Developer platformsCoding
   Open sourceVisit Repomix (official site)

CodeGraph FAQ

Sources & verification

Facts on this page come from the official sources below, each with the date we read it. Written by Debuly; the vendor did not review this page.

Image credits

• Logo: none collected yet, so a monogram is shown.
• Cover: none collected yet, so a generated tile is shown.

Spot an error? Report an issue with this listing. We also remove a logo or image on the owner's request.