reviewing-agent-prompting

Audit and improve coding agent prompts using state-machine structure.

Updated Dec 30, 2025

Search Tags:#governance #instruction #system-prompt #state-machine #agent-prompt #prompt-review #prompt-audit

Ready to run

npx skills add https://github.com/mikekelly/reviewing-agent-prompting --skill reviewing-agent-prompting

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: reviewing-agent-prompting
Source: https://github.com/mikekelly/reviewing-agent-prompting/tree/main
Command: npx skills add https://github.com/mikekelly/reviewing-agent-prompting --skill reviewing-agent-prompting

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill helps teams audit and improve prompts for coding agents by applying state-machine thinking to impose structure, reduce hallucinations, and improve reliability.

Core Features & Use Cases

State-machine framing for prompts (entry conditions, actions, exit criteria, transitions).
Structured review framework (phase gates, output requirements, escalation, anti-patterns).
Guidance for auditing system prompts, task delegation prompts, or instruction prompts in software engineering workflows.
Quick-start guidance to obtain a structured evaluation and recommendations.

Quick Start

Provide the prompt text you want analyzed to receive a structured, gate-aware evaluation.

Frequently Asked Questions about reviewing-agent-prompting

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

FAQPage Schema

How do I audit and improve prompts for coding agents to reduce hallucinations?▼

To audit and improve prompts for coding agents, you apply state-machine structure to enforce clear entry conditions, actions, exits, and transitions. This reduces hallucinations by imposing strict phase gates and explicit output formats.

What is state-machine framing for agent prompt review?▼

State-machine framing for agent prompt review structures prompts into explicit states with entry conditions, defined actions, exit criteria, and transitions. This enforces clear phase gates and escalation triggers to improve reliability.

How do I structure system prompts with phase gates and exit criteria?▼

You structure system prompts by applying a structured review framework that defines phase gates, explicit output requirements, escalation triggers, and anti-patterns. This ensures first-action clarity and reliable state transitions.

Can I use this prompt audit framework for task delegation and instruction prompts?▼

Yes, the prompt audit framework is applicable to system prompts, task delegation prompts, and instruction prompts across software engineering workflows. It evaluates and refines all these prompt types using state-machine logic.

What is the best way to evaluate prompt anti-patterns and escalation triggers?▼

The best way to evaluate prompt anti-patterns and escalation triggers is through a structured review framework. It audits prompts against defined anti-patterns and ensures escalation triggers are explicitly stated for first-action clarity.