reviewer-agent

Review bounded work for correctness and output evidence-linked findings with a disposition.

Updated Mar 30, 2026

Search Tags:#audit #evidence #review #code-review #read-only #disposition

Ready to run

npx skills add https://github.com/harkers/forge-email-server --skill reviewer-agent-harkers

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: reviewer-agent
Source: https://github.com/harkers/forge-email-server/tree/main/skills/reviewer-agent
Command: npx skills add https://github.com/harkers/forge-email-server --skill reviewer-agent-harkers

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This skill enables a structured, independent read-only review of bounded work, ensuring findings are tied to concrete evidence and accompanied by a clear disposition recommendation.

Core Features & Use Cases

Read-only assessment of correctness gaps without becoming the implementer.
Tie findings to specific evidence and relevant files or behaviors.
Produce a concrete disposition: accept, revise, reject, or escalate for timely action.

Quick Start

Provide the artifact to be reviewed and request an independent, evidence-based assessment with a recommended disposition.

Frequently Asked Questions about reviewer-agent

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

FAQPage Schema

How do I perform an independent code review with evidence-backed findings?▼

An independent code review with evidence-backed findings assesses bounded work in read-only mode, tying correctness gaps to specific files or behaviors to produce a structured list of findings with severities and evidence references.

What is an evidence-based review disposition and when is it needed?▼

An evidence-based review disposition is a recommendation to accept, revise, reject, or escalate work after implementation, documentation, deployment, or investigations when an independent assessment pass is required for correctness.

How do I get a structured severity list for my code review findings?▼

To get a structured severity list for code review findings, provide the completed artifact for a read-only assessment that ties each finding to concrete evidence and outputs an acceptance recommendation.

Can I use a read-only review process without modifying the implementation?▼

Yes, a read-only review process assesses correctness gaps without becoming the implementer, ensuring the review remains independent and only outputs findings tied to evidence with a disposition recommendation.

Does an independent review pass work for post-deployment and investigation artifacts?▼

Yes, an independent review pass works for post-deployment and investigation artifacts, applying a read-only assessment to determine an accept, revise, reject, or escalate disposition based on evidence.

What is the best way to audit bounded work for correctness gaps?▼

The best way to audit bounded work for correctness gaps is an independent, read-only review that ties findings to concrete evidence and produces a clear disposition recommendation for timely action.