tirth8205/code-review-graph v2.1.0

v2.1.0 — Notebooks, Lua, Zero-Config Onboarding

on GitHub

What's New

Jupyter & Databricks Notebook Parsing

Parse .ipynb files and Databricks .py notebook exports. Extracts functions, classes, and imports across Python, R, and SQL cells — your notebooks are now part of the graph.

Lua Language Support (20th language)

Full parsing for .lua files: functions, local functions, method calls, and require statements.

Perl XS Support

Parse .xs files with improved Perl call detection and test coverage.

Zero-Config Onboarding

install now sets up skills, hooks, and CLAUDE.md by default. Graph instructions are injected into all platform rule files automatically. Smart detection of uvx vs pip install method.

JS/TS Arrow Function Indexing

const foo = () => {} and const bar = function() {} now correctly appear as nodes in the graph.

Website Launch

code-review-graph.com is live.

All Changes

Added

• Jupyter notebook parsing (.ipynb) with multi-language cell support (Python, R, SQL)
• Databricks .py notebook export parsing with # COMMAND ---------- cell boundaries
• Lua language support — functions, local functions, method calls, requires
• Perl .xs file support with improved call detection
• Zero-config onboarding: skills, hooks, and CLAUDE.md installed by default
• Platform rule injection into all supported platform config files on install
• Smart install method detection (uvx vs pip/pipx)
• --platform claude-code alias

Fixed

• JS/TS arrow functions and const exports now indexed as nodes (#66)
• importers_of path normalization with resolve() (#65)
• Custom embedding model architecture support restored (#79)

Docs

• README overhauled with all 9 diagrams, website link, platform-neutral language

Full Changelog: v2.0.0...v2.1.0

pip install --upgrade code-review-graph