Codebase context
codebase indexing, repo maps and code search for agents
Graphify-Labs/graphify Graphify-Labs/graphify is the clear leader, offering a widely adopted, deterministic AST-based repo-wide knowledge graph with EXTRACTED/INFERRED edge labels. Before choosing, decide whether you need that single, local-first graph model or a different tradeoff (repo-embedded docs, hybrid vector/BM25 retrieval, lightweight local search, or richer type/dataflow analyses).
Which one matches your setup?
Answer any of the questions — the shortlist updates as you go. Recommendations come from the capability passports below, nothing else.
Comparison matrix
Axes are extracted from each project's docs by our review pipeline; the maturity score is computed from stars, growth and commit activity — not an opinion. Click a column to sort.
| Runs in | Models | Context | Cost to run | |||||
|---|---|---|---|---|---|---|---|---|
⭐ 114.6k +2819/7d | CLICoding-agent pluginIDECI | BYOKOpenAIAnthropicGeminiOllama / localFixed provider | Whole-repo analysis | Edge confidence tags | ●●●●● | Fully local possible | ●●●●● | Free for code (local); optional API key for semantic media/docs |
⭐ 5.5k +498/7d | CLIIDECoding-agent pluginCI | BYOKOpenAIAnthropic | Whole-repo graph | Ranking & checks | ●●●●● | Self-hostable | ●●●●● | Your API key; provider billing applies. |
⭐ 12.5k +14/7d | CLIIDECoding-agent plugin | BYOKOpenAI | Whole-repo retrieval | Hybrid search + filters | ●●●●● | Cloud APIs (your keys) | ●●●●● | Requires OpenAI and Zilliz Cloud (Milvus) API keys. |
⭐ 2.7k +16/7d | CLICoding-agent plugin | BYOKOllama / local | Whole-repo analysis | Path/lang filters | ●●●●● | Fully local run | ●●●●● | Local embeddings free with the 'full' variant; slim/latest uses cloud embeddings and your API key. |
⭐ 5.5k +498/7d | CLICoding-agent pluginIDE | BYOKOpenAIAnthropicOllama / localFixed provider | Whole-repo analysis | Minimal filters | ●●●●● | Fully local possible | ●●●●● | Your API key, per-use |
⭐ 1.2k +10/7d | CLICI | AnthropicOpenAI | Diff + related files | Authoritative surface | ●●●●● | Fully local | ●●●●● | Free, local binary |
⭐ 765 | CLICoding-agent plugin | BYOK | Whole-repo index | Reranker + RRF | ●●●●● | Cloud APIs (your key) | ●●●●● | Requires your API key for embedding and reranker calls |
⭐ 213 +1/7d | CLICICoding-agent pluginIDE | BYOK | Whole-repo analysis | Configurable filters | ●●●●● | Fully local | ●●●●● | Free, local (no hosted service) |
⭐ 873 +4/7d | CLICoding-agent plugin | BYOK | Whole-repo analysis | Truncation & tiers | ●●●●● | Self-hostable | ●●●●● | Self-hosted; external model/API costs apply |
⭐ 458 | CLICoding-agent plugin | Fixed provider | Whole-repo analysis | Quality rules & validate | ●●●●● | Self-hostable | ●●●●● | Free, local |
⭐ 156 | CLICIGitHub ActionIDECoding-agent plugin | BYOK | Diff + related files | Rule IDs + lint | ●●●●● | Fully local | ●●●●● | Free, open-source |
Capability profiles
Six axes, 0–5 each. The shape tells you the strategy: a wide hexagon is a generalist, a spike is a specialist. Showing the 8 most established — the rest are in the full catalog.
Builds a deterministic, AST-based, repo-wide knowledge graph (no embeddings or vector store) with every edge labeled EXTRACTED or INFERRED.
Persists a deterministic, human-readable context graph as markdown files inside the repo (no embeddings or external index), so agents can follow real files and share an up-to-date code understanding via git.
An MCP plugin that gives Claude Code whole-repo semantic code search via hybrid BM25 + dense vector retrieval backed by Milvus, reducing tokens compared with loading large directories into the model.
AST-based semantic search that is extremely lightweight, agent-friendly (skill/MCP plugin), and can run fully locally with offline embeddings.
Writes a readable, linked markdown graph of your code (no embeddings or external index), letting agents use a local, regenerable map rather than re-exploring the repo each run.
Makes the repository the default system of record for agent context and durable execution plans, keeping intent, tests, and validation in Git rather than in a hidden workflow database.
Native Rust CLI that combines AST-based call-graph analysis with a hybrid dense+sparse+RRF+cross-encoder reranker search pipeline and MCP integration for Claude Code.
Provides persisted, repository-level structural analyses and composed agent workflows (plan/review/context) via a local MCP server, beyond editor-focused tools.
All repositories (11)
Turn any codebase, with its docs, SQL schemas, configs, and PDFs, into a queryable knowledge graph. A /graphif…
Code search MCP for Claude Code. Make entire codebase the context for any coding agent.
Turbocharge Claude Code, Cursor, Codex, Gemini & every coding agent: faster, cheaper, with contextual understa…
Turbocharge Claude Code, Cursor, Codex, Gemini & every coding agent: faster, cheaper, with contextual understa…
A super light-weight embedded code search engine CLI (AST based) that just works - improves speed and efficie…
Turn any repo into an agent-ready workspace for Claude Code, Codex, Cursor, and other coding agents.
100% Rust implementation of code graphRAG with blazing fast AST+FastML parsing, surrealDB backend and advanced…
Rust-powered code intelligence CLI for AI coding agents. Builds call graphs and hybrid semantic search indexes…
Living project docs for coding agents: keep guides, progress logs, change maps, and handoff context updated as…
Local-only Go static analysis engine with a built-in MCP server. Gives AI coding agents deterministic structur…
Open, MIT context format for AI coding agents: centralized .aigx/ rules + a per-file boundary index that targe…
Hosted alternatives
If running your own reviewer is more ops than you want, these managed services cover the same job.
Managed code search and codebase context across your repositories, so agents and developers query a hosted index instead of one you build and keep in sync.
Try Sourcegraph →Serves full-codebase context as an API for AI tools: point at your repos and get cross-file understanding hosted rather than running your own code-graph.
Try Greptile →Managed Milvus from the Zilliz team, the hosted vector store behind many code-context setups, run for you instead of operated alongside your indexer.
Try Zilliz Cloud →