gitcrawl
A local-first GitHub triage tool for maintainers and agents. Sync issues and PRs into SQLite for search, clustering, and terminal review; use Octopool for the pooled gh cache that used to live here.
#Two jobs, one binary
gitcrawl mirrors a GitHub repository's issues and pull requests into local SQLite, then layers semantic clustering and full-text search on top so a maintainer (or an agent acting on their behalf) can triage threads without burning live search quota.
- Local SQLite first. All issues, PRs, comments, reviews, files, commits, checks, and workflow runs land in
~/.config/gitcrawl/gitcrawl.db. Queries hit the disk, not GitHub. - Octopool for
gh. The oldgitcrawl ghcache moved to Octopool, which owns the shared org-authenticated GitHub relay. - Semantic clustering. OpenAI embeddings group related reports, with deterministic GitHub reference evidence (
#123,pull/123) preventing weak similarity bridges from forming mega-clusters. - Terminal UI.
gitcrawl tuiis a keyboard- and mouse-driven cluster browser with bidirectional sort, jump-to-number, neighbors, and member-level governance actions. - Agent-friendly JSON. Every command supports
--jsonfor clean automation surfaces.
#Pick your path
#I want to try it
Quickstart walks you from git clone to a populated cluster view in five minutes.
#I want to wire up an agent
Use gitcrawl search for local discovery and Octopool migration for pooled gh reads.
#I want to triage a busy repo
Read Sync to bring data local, then Clustering and the TUI for the maintainer workflow.
#I want the full reference
Commands lists every flag and JSON field. Configuration covers env vars and paths.
#Project status
Early bootstrap. The implementation is being built in small commits — see the changelog for what shipped recently.
The product contract in SPEC.md is the source of truth for what is in and out of scope.
#Out of scope
- Local HTTP API
- Hosted service runtime
- Browser web UI
- GitHub write-back actions (use
ghfor those)
#License
Released under the MIT license.