harness-review

Automate multi-agent code review workflows and aggregate verdicts.

3|1|Updated May 1, 2026

Search Tags:#automation #code review #parallel processing #quality gate #verdict

Ready to run

npx skills add https://github.com/bigbulgogiburger/claude_jira_harness --skill harness-review-bigbulgogiburger

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: harness-review
Source: https://github.com/bigbulgogiburger/claude_jira_harness/tree/main/skills/harness-review
Command: npx skills add https://github.com/bigbulgogiburger/claude_jira_harness --skill harness-review-bigbulgogiburger

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill automates the review of code changes across multiple projects by deploying specialized agents to evaluate quality and generate a comprehensive verdict, reducing manual review efforts.

Core Features & Use Cases

Parallel Multimodal Review: Fan-out code analysis to project-specific agents for detailed assessment.
Verdict Aggregation: Compile individual agent outputs into a final PASS, ITERATE, or ESCALATE decision.
Use Case: For a large codebase update, instantly evaluate all impacted modules and determine if code is ready for merge or needs further revisions.

Quick Start

Invoke the harness-review skill to analyze recent changes and obtain an overall quality verdict.

Frequently Asked Questions about harness-review

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

FAQPage Schema

How do I automate code reviews across multiple projects?▼

Automated code review uses parallel fan-out analysis to deploy specialized agents across impacted modules, aggregating individual assessments into a conclusive PASS or ITERATE verdict.

What is parallel fan-out in automated code review?▼

Parallel fan-out in code review dispatches project-specific agents simultaneously to evaluate distinct codebase modules, compiling their individual outputs into a final aggregated quality verdict.

How do I get a final merge decision from automated code analysis?▼

Automated code analysis aggregates multi-agent review outputs into a conclusive verdict, determining whether code changes are ready to PASS, require ITERATE revisions, or need ESCALATE review.

Can I use multi-agent code review for large codebase updates?▼

Multi-agent code review suits large codebase updates by fanning out parallel analysis to assess all impacted modules and instantly determining if code meets quality standards for merge.

What are the limitations of automated quality gates in code review?▼

Automated quality gates rely on specialized agent outputs, meaning complex architectural changes might produce ITERATE or ESCALATE verdicts requiring manual intervention rather than a direct PASS decision.