code-review-graph vs openqodex
Local-first code intelligence graph for AI coding tools: Tree-sitter AST graph + blast-radius analysis served over MCP, so agents read ~82x fewer tokens per review question. — versus — Local pre-push code review for coding agents: deterministic scanners (gitleaks, semgrep) plus separate Claude Code/Codex reviewer sessions, with a push gate. No account or key of its own.
Both give reviewers a code graph to judge blast radius. code-review-graph is the graph layer; OpenQodex is the full pre-push review loop.
| code-review-graph | openqodex | |
|---|---|---|
| Stars | 33k | 1.1k |
| Forks | 3.0k | 76 |
| Language | Python | TypeScript |
| License | MIT | Apache-2.0 |
| Last activity | 8 days ago | yesterday |
| Topics | code-intel | coding, evals |
| Curated connections | 9 | 3 |
code-review-graph — the curator's take
The rare benchmark-honest repo: it tells you the 528x number is the best case and its recall metric is circular. Install once, it configures 14 platforms (Codex, Claude Code, Cursor...). Reach for it on monorepos where review context is the token sink; skip on small repos where a grep costs less than the graph's own metadata.
openqodex — the curator's take
Use it when agents write most of your code and you want a second, independent reviewer before every push, running on your existing Claude Code or Codex login. The writer agent never reviews its own change by default. Skip it on Windows (untested), or if you already run review in CI on PRs and don't want reviews to cost tokens locally. Cursor and Cline can trigger a review but can't be the reviewer.