Alibaba open-sources its battle-tested code reviewer: 7%→34% precision, 93% fewer tokens
victor_explore · x · 2026-11-15
Alibaba open-sourced open-code-review, a hybrid code review tool it ran internally for two years: deterministic pipelines pick files and apply rules while the LLM agent only judges. Same Opus 4.6 jumps from 7% to 34% precision and drops from 5.6M to 385K tokens. The repo (51.6K stars, Apache-2.0) ships multi-language rulesets (NPE, thread-safety, XSS, SQL injection) and is OpenAI/Anthropic compatible. Its benchmark shows Claude-4.6-Opus hitting F1 25.10% on this harness vs 11.57% under Claude Code, underscoring how much harness design matters.
More from coding & agent
- Daptin Adds OpenAI-Compatible LLM Routing With Budgets, Limits and Streaming — artpar · 2026-11-15
- Blender MCP hits 34.5k stars: direct 3D modeling in plain language with any LLM — sidahuj · 2026-11-15
- 18 agents coding overnight on one Linux box: 'I'm the bottleneck now' — AIOSAI · 2026-11-15
- IBM Engineer on Trustworthy AI Evals: Calibrate LLM Judges, Check Agent Trajectories — TejasKumar_ · 2026-11-15
- Antigravity Browser Agent Swarm Turns a Portrait Into Google Sheets Pixel Art — Saboo_Shubham_ · 2026-11-15
- CrommaCode launches: a gamified desktop terminal for watching Claude Code and Codex work — huangyun_122 · 2026-11-15