What problem does it solve?
Reviewers often struggle to quickly understand what a pull request or local branch actually changed, whether the changes exceed the stated requirement, and whether the code matches project conventions. This Skill reconstructs the requirement and implementation boundaries from the real diff, then flags scope creep, quality risks, and style inconsistencies.
Core Features & Use Cases
- Requirement Reconstruction: Reads the actual diff against upstream/main to explain what changed, why it matters, and which modules are affected, rather than just restating the PR title.
- Scope & Quality Review: Classifies changes as reasonable expansion, suspicious overreach, or high-risk overreach, and identifies missing tests, hardcoded logic, and untestable code placement.
- Style Consistency Check: Compares changes against neighboring implementations, including FastGPT-specific rules like static i18n keys and parseApiInput contract boundaries.
- Use Case: A reviewer receives a large FastGPT PR mixing a feature with unrelated refactoring. Invoke this Skill to get a structured Chinese report separating real behavior changes from noise, with file paths and line numbers for each issue.
Quick Start
Ask the assistant to use $pr-change-analysis to analyze a specific PR URL or the current local branch against upstream/main and produce a structured review report.