# trailhq/Graft

Turbocharge Claude Code, Cursor, Codex, Gemini & every coding agent: faster, cheaper, with contextual understanding specific to your codebase.

Repository: https://github.com/trailhq/Graft
Canonical: https://ross.abutalabs.com/products/graft
Homepage: https://graft.nanonets.ai
Language: TypeScript
License: MIT
License Family: permissive
Topics: ai-agents, claude-code, code-graph, context-engineering, developer-tools, knowledge-graph, llm, mcp, mcp-server, tree-sitter, anthropic, cli, codex, cursor, gemini, open-source, openai
Last push: 2026-08-25T20:18:35+00:00

## Health v2 (maintenance only)
Score: 58/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 99, release rhythm 35, longevity 4
- inputs: {"age_days": 61, "days_push": 8, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, young
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 4948, forks 439 (observed 2026-08-28T04:09:02.609728+00:00)

## What it is
Graft is an open-source context layer for large codebases that builds a deterministic code graph using tree-sitter AST parsing and exposes it via an MCP server and CLI. It integrates with coding agents like Claude Code, Cursor, Codex, and Gemini to reduce token usage, latency, and cost while improving correctness.

## Use cases
- give claude code contextual understanding of my codebase
- reduce token costs when using coding agents on large repos
- find the blast radius of a symbol before refactoring
- speed up agent cold starts instead of grepping from scratch
- run an mcp server that maps my repository structure
- make cursor or codex cheaper and faster on my project

## When to choose
- you use AI coding agents on large codebases and want lower cost and latency
- you want a fully local, deterministic code context layer without vector embeddings
- you need impact analysis across 20+ languages via tree-sitter

## When to avoid
- you need semantic/embedding-based code search rather than structural graph context
- your project is small enough that agents already work fine without extra context
- you need an IDE plugin with a GUI rather than a CLI/MCP integration

## Facets
- artifact type: cli-tool
- maturity: active
- function: mcp, parser, developer-tools, search-engine, cli
- domain: developer-tools, large-language-models, parsers, apis
- platform: cli, cross-platform, windows
- tags: code-graph, context-engineering, knowledge-graph, tree-sitter, claude-code, cursor, codex, gemini, coding-agents, impact-analysis, mcp-server, local-first, ai-agents, nodejs, macos, linux

## Member repositories
- trailhq/Graft (main) score 58

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:09:02.609728+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T18:17:52.911579+00:00, confidence not recorded.
  - readme: https://github.com/trailhq/Graft (fetched 2026-08-28T04:09:02.609728+00:00, sha 2f1fbf43ffd5)
  - homepage: https://graft.nanonets.ai (fetched 2026-08-29T08:59:43.437019+00:00, sha 6bfdb893ead3)
- Data as of 2026-08-30T08:39:29.467469+00:00.