diff-eval

Community

Score agent PR changes with structured metrics.

Authorwzh4464
Version1.0.0
Installs0

System Documentation

What problem does it solve?

Evaluates agent-generated code changes against human-approved PRs using a formal rubric, diff analysis, and auto-generated file filtering to ensure alignment with ground truth.

Core Features & Use Cases

  • Structured diff comparison: compare local agent changes (relative to PR base) with the ground-truth PR diff, while filtering out auto-generated files.
  • Dual analysis: perform data-based coverage on file-level changes and semantic checks on functional intent.
  • Scoring and reporting: produce a formal evaluation report based on functional correctness, completeness, and behavioral equivalence.
  • Use cases: pre-merge validation of AI-generated patches, benchmarking code generation tools, and artifacting evaluation results for audits.

Quick Start

Compute the base commit, derive the generated patch, compare it against the PR diff, and generate a full evaluation report.

Dependency Matrix

Required Modules

None required

Components

references

💻 Claude Code Installation

Recommended: Let Claude install automatically. Simply copy and paste the text below to Claude Code.

Please help me install this Skill:
Name: diff-eval
Download link: https://github.com/wzh4464/claude-skills/archive/main.zip#diff-eval

Please download this .zip file, extract it, and install it in the .claude/skills/ directory.
View Source Repository

Agent Skills Search Helper

Install a tiny helper to your Agent, search and equip skill from 471,000+ vetted skills library on demand.