audit-instructions

Audit agent instruction files against source-of-truth code with a five-pass procedure.

Updated Mar 3, 2026
One-click install
npx skills add https://github.com/0xb11a/crypto-claw --skill audit-instructions
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: audit-instructions
Source: https://github.com/0xb11a/crypto-claw/tree/main/.claude/skills/audit-instructions
Command: npx skills add https://github.com/0xb11a/crypto-claw --skill audit-instructions

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Audits CryptoClaw's agent instruction files for consistency, correctness, and completeness against source-of-truth code to prevent drift and runtime issues.

Core Features & Use Cases

  • Systematic 5-pass audit procedure (command validity, script references, constant consistency, cross-file behavior, completeness)
  • Generates structured, prioritized reports with clear severities (CRITICAL, WARNING, INFO)
  • Applicable to multi-agent instruction ecosystems to ensure alignment between docs and implementation

Quick Start

Run the audit against all targeted agent instruction files to generate a comprehensive report.

Frequently Asked Questions about audit-instructions

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

FAQPage Schema
How do I audit agent instruction files for consistency against source code?

Multi-agent instruction ecosystems require consistency audits to prevent drift between documentation and implementation. A five-pass audit procedure verifies command validity, script references, constant consistency, cross-file behavior, and completeness to ensure alignment across the ecosystem.

What is the best way to check multi-agent documentation for correctness and completeness?

A five-pass audit procedure checks multi-agent documentation against source-of-truth code to ensure correctness and completeness. It validates command validity, script references, constant consistency, cross-file behavior, and completeness, outputting a single structured report with severity tagging.

How do I generate a structured audit report with severity tagging for instruction files?

Running a consistency audit across targeted instruction files generates a structured report with severity tagging. The audit classifies issues as CRITICAL, WARNING, or INFO based on alignment with source-of-truth code, preventing runtime failures.

Can I use an automated audit to prevent drift in multi-agent instruction ecosystems?

An automated audit prevents drift in multi-agent instruction ecosystems by enforcing alignment across commands, scripts, constants, and cross-behavior rules. It checks instruction files against source-of-truth code to prevent runtime failures and generate structured reports.

What does the five-pass audit procedure check for when reviewing agent instructions?

When reviewing agent instructions, the five-pass audit procedure checks command validity, script references, constant consistency, cross-file behavior, and completeness. It ensures documentation aligns with source-of-truth code to prevent drift and runtime failures.

When should I run a consistency audit on my agent documentation?

Run a consistency audit on agent documentation when alignment with source-of-truth code must be verified to prevent drift. It is essential for multi-agent instruction ecosystems to ensure correctness, completeness, and prevent runtime failures.