From 8b2652704db17fce22a0b044a48a362927295ddc Mon Sep 17 00:00:00 2001 From: best Date: Sun, 22 Feb 2026 21:44:08 +0800 Subject: [PATCH] chore: clarify ai-review workflow is static rule-based, not LLM --- .github/workflows/ai-review.yml | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/.github/workflows/ai-review.yml b/.github/workflows/ai-review.yml index 72b7db9..9ff362f 100644 --- a/.github/workflows/ai-review.yml +++ b/.github/workflows/ai-review.yml @@ -1,5 +1,11 @@ name: AI Code Review +# NOTE: This workflow performs STATIC RULE-BASED checks (pattern matching on diff), +# NOT true AI semantic analysis. It detects common issues such as `any` usage, +# empty catch blocks, missing test files, and exported symbol changes. +# For AI-powered semantic review, the dev-mgr agent performs manual review +# during patrol cycles using `gh pr diff` + LLM analysis. + on: pull_request: types: [opened, synchronize]