review-contract

Standardize code review findings into a canonical JSON schema.

3|Updated May 12, 2026

Search Tags:#automation #developer tools #quality assurance #static analysis #code review #json

Ready to run

npx skills add https://github.com/gao-hongnan/omniagents --skill review-contract-gao-hongnan

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: review-contract
Source: https://github.com/gao-hongnan/omniagents/tree/main/plugins/reviewer/skills/review-contract
Command: npx skills add https://github.com/gao-hongnan/omniagents --skill review-contract-gao-hongnan

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

This Skill eliminates inconsistent, subjective, and un-actionable code review feedback by enforcing a strict, machine-readable JSON contract for all review findings.

Core Features & Use Cases

Structured Findings: Enforces a canonical JSON schema for all review findings, ensuring every report includes required fields like file, line, severity, and blast radius.
Automated Verification: Provides a verifier agent that aggregates findings, deduplicates issues, and adjudicates severity based on a predefined rubric.
Use Case: Use this when you need to scale code reviews across multiple specialist agents (security, performance, correctness) while ensuring the final report is consistent, jump-to-able, and free of noise.

Quick Start

Run the review-contract skill to initiate a structured code review of the current changes and generate a consolidated JSON report.

Frequently Asked Questions about review-contract

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I standardize automated code review findings into a consistent JSON format?▼

To standardize automated code review findings into a consistent JSON format, enforce a strict canonical schema that captures file, line, severity, and blast radius. This ensures deduplication, cross-specialist aggregation, and machine-readability across all specialist agents.

What is the best way to deduplicate and aggregate code review findings from multiple specialist agents?▼

The best way to deduplicate and aggregate code review findings from multiple specialist agents is to use a verifier agent that applies a predefined rubric to adjudicate severity, consolidating cross-specialist reports into a single, consistent, machine-readable JSON output.

How does structured static analysis enforce severity and blast radius in PR review workflows?▼

Structured static analysis enforces severity and blast radius in PR review workflows by applying strict schema validation to every finding. This guarantees that required fields like file citations and impact analysis are present, ensuring reports are jump-to-able and actionable.

Can I apply strict schema validation to scale code reviews across security and performance agents?▼

Yes, you can scale code reviews across security and performance agents by enforcing a strict JSON contract. This approach eliminates subjective feedback, ensures machine-readability, and provides a verifier agent to adjudicate severity and deduplicate cross-specialist issues.

When do I need a canonical JSON contract for automated code review reports?▼

You need a canonical JSON contract for automated code review reports when you want to eliminate inconsistent, un-actionable feedback and ensure every issue includes jump-to-able file citations, blast radius analysis, and strict schema validation for downstream processing.