Practical overview
When it fits
Use Graft when coding-agent sessions repeatedly rediscover repository structure, conventions or relationships, and a local CLI-based context layer is appropriate.
Who it is for
- Developers using coding agents on large or repeatedly explored codebases.
- Teams that want repository-specific context and shared agent wiring.
- Users of Claude Code, Cursor, Codex or Gemini workflows.
Documented uses
- Build a local knowledge or code graph from a repository.
- Provide coding agents with codebase-specific context.
- Reduce repeated exploration, tool calls, tokens and execution time.
- Carry project rules and corrections into later agent sessions.
- Share agent wiring while allowing each teammate to generate a local graph.
Requirements & limits
The repository describes Graft as open source and MIT-licensed. It is distributed as an npm CLI and can also be invoked with npx.
- The documented setup is CLI-based and uses npm or npx.
- Named integrations include Claude Code, Cursor, Codex and Gemini.
- The graph is generated locally and added to .gitignore rather than committed.
- The repository describes background rebuilding after agent turns and no daemon by default.
- The project is written in TypeScript and the repository is not archived.
Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.
Set up Graft with a coding-agent workflow
Build a local codebase graph and wire it into a selected supported coding agent.
Before you start
- npm or npx access
- A codebase to index
- A supported coding-agent workflow: Claude Code, Cursor, Codex or Gemini
Steps
Install the CLI with npm install -g @nanonets/graft, or use npx @nanonets/graft.
Run graft init.
Choose the coding agent to wire up, or use graft init --agents claude for Claude Code alone.
Review planned file changes first with graft init --dry-run when needed.
Commit the generated .claude/ wiring if sharing the setup with teammates; each teammate runs graft build to create a local graph.
Expected outcome
A local graft/ graph is created and the selected agent workflow is wired to use repository context.
Guide limitations
- The documented evidence does not provide agent-specific configuration details beyond the named commands and choices.
- The graph is a local regenerable cache and is not intended to be committed.
Compare options for AI coding assistance
These projects document a common task and share a project role. This is a source-based comparison, not an execution benchmark or a guarantee of interchangeability.
| Project | Documented purpose | Platform | Access and requirements |
|---|---|---|---|
| Graft | Build a local knowledge or code graph from a repository. | CLI | The repository describes Graft as open source and MIT-licensed. It is distributed as an npm CLI and can also be invoked with npx. No pricing or account requirement is documented in the supplied evidence. |
| Continuum — pouyahasanamreji | Store project lessons so coding agents can retrieve them across conversations. | CLI, Self-hosted, Server | The repository documents local Docker Compose and bare-metal development setups. Bare-metal startup requires Node 22.12.0 or later and pnpm 10.33 or later; the Compose path skips these host prerequisites. |
| Haft | Record the problem a project is solving and the options considered. | CLI, Self-hosted |
Sources & review
Source record details
- Source availability
- Active
- Sources checked
- 1 October 2026
- Original website title
- trailhq/Graft · Official homepage title · retained in its source language