audit-loop-run

Audit long-horizon AI loop execution history and state transitions.

7|2|Updated Jan 3, 2026
One-click install
npx skills add https://github.com/BrennonTWilliams/little-loops --skill audit-loop-run
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: audit-loop-run
Source: https://github.com/BrennonTWilliams/little-loops/tree/main/skills/audit-loop-run
Command: npx skills add https://github.com/BrennonTWilliams/little-loops --skill audit-loop-run

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill addresses the challenge of verifying whether an AI-driven loop actually achieved its intended goal, helping to identify phantom successes, structural defects, and inefficient iterations.

Core Features & Use Cases

  • Goal Verification: Audits loop execution against stated goals, checking for artifact mutations and threshold contract satisfaction.
  • Fault Detection: Identifies specific failure signals like evaluate errors, throttle stops, and over-escaped shell corruption.
  • Improvement Proposals: Generates ranked, actionable recommendations to refine loop structure and evaluator logic.

Quick Start

Use the audit-loop-run skill to assess the effectiveness of the most recent execution of the loop named apo-textgrad.

Frequently Asked Questions about audit-loop-run

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

FAQPage Schema
How do I verify that an AI loop actually achieved its intended goal?

To verify AI loop effectiveness, you audit execution history and state transitions against stated goals, checking for artifact mutations and threshold contract satisfaction to detect phantom successes.

What is a phantom success in automated software development workflows?

A phantom success occurs when a long-horizon AI loop reports goal achievement without actually satisfying threshold contracts or artifact mutations, which auditing event logs can detect and flag.

How do I identify performance bottlenecks and structural failures in long-horizon AI loops?

Identify performance bottlenecks and structural failures in long-horizon AI loops by analyzing event logs, artifact mutations, and state transitions to detect evaluate errors, throttle stops, and shell corruption.

Can I generate improvement proposals for an automated loop that failed its quality assurance audit?

Yes, auditing loop execution generates ranked, actionable improvement proposals that refine loop structure and evaluator logic, providing scorecard reporting for automated software development workflows.

What specific failure signals should I look for when debugging state transitions in an FSM loop?

When debugging FSM loop state transitions, look for specific failure signals like evaluate errors, throttle stops, and over-escaped shell corruption by analyzing event logs and artifact mutations.