claude-context vs Graft
claude-context is much bigger: 12.5k stars against 5.6k.
Graft leads on privacy and setup ease. claude-context does not take any axis by a clear margin.
Stars and commit dates come from our own daily tracking. The six axes are read off each project's documentation by our review pipeline, so they describe what a project says about itself, not what we measured in its code.
Where they stand today
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.
- Stars
- 12.5k
- Tracked growth
- +6%
- Maturity
- ●●●●●
- Last commit
- 52d ago
- Language
- TypeScript
- License
- MIT
- Cost to run
- Requires OpenAI and Zilliz Cloud (Milvus) API keys.
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.
- Stars
- 5.6k
- Tracked growth
- not tracked long enough
- Maturity
- ●●●●●
- Last commit
- 1d ago
- Language
- TypeScript
- License
- MIT
- Cost to run
- Your API key, per-use
Six axes, head to head
Each axis runs 0 to 5. The label under a score is what that project's own docs claim, not a category average.
| Axis | claude-context | Graft |
|---|---|---|
Context depth How much of your codebase it sees before it answers: the open diff, the diff plus related files, or the whole repository. | ●●●●● Whole-repo retrieval | ●●●●● Whole-repo analysis |
Noise control How it keeps output volume down — severity thresholds, deduplication, incremental runs over new commits only. | ●●●●● Hybrid search + filters | ●●●●● Minimal filters |
Customization How far it bends to your team: custom rules, prompts, style guides, per-path config. | ●●●●● Configurable rules | ●●●●● Config options |
Privacy Whether your code stays on your own infrastructure: fully local, self-hostable, or cloud API only. | ●●●●● Cloud APIs (your keys) | ●●●●● Fully local possible |
Model freedom Whether you can point it at any provider, or it is wired to one. | ●●●●● Bring-your-own-key | ●●●●● Bring-your-own-model |
Setup ease What it takes to get a first useful run out of it. | ●●●●● API key + config | ●●●●● Two-command setup |
Which one to pick
Pick claude-context if…
Cost-effective — provides whole-repo semantic retrieval (hybrid BM25 + dense embeddings) as an MCP server to give deep context to Claude Code while lowering token usage.
Pick Graft if…
Efficiency-first — pick Graft when you want much faster and cheaper agent runs (often with equal or better correctness) by providing a repo-wide local node graph that agents can read directly.
- Privacy: Fully local possible (5/5 against 3/5)
- Setup ease: Two-command setup (5/5 against 3/5)
What people want from each one
zilliztech/claude-context
trailhq/Graft
Questions people ask
Is claude-context better than Graft?
Graft leads on privacy and setup ease. claude-context does not take any axis by a clear margin. claude-context is worth picking when cost-effective — provides whole-repo semantic retrieval (hybrid BM25 + dense embeddings) as an MCP server to give deep context to Claude Code while lowering token usage.
Which of claude-context and Graft keeps my code private?
claude-context: Cloud APIs (your keys) (3/5). Graft: Fully local possible (5/5).
What does each one cost to run?
claude-context: Requires OpenAI and Zilliz Cloud (Milvus) API keys.. Graft: Your API key, per-use.
Full profiles: zilliztech/claude-context and trailhq/Graft. Everything else in Codebase context.