Universal Context Auditor

The "Context Auditor" - evaluates AI configuration files against research-backed principles to maximize task success rate while minimizing cost and attention decay.

What This Does

Audits your existing AI configuration files (CLAUDE.md, .cursorrules, instructions.md, etc.) and provides:

• Structural analysis against LLM research (Stanford, UC Berkeley, Anthropic)
• Fidelity scoring - Will the AI reliably follow your rules?
• Cache efficiency - Is your config optimized for prompt caching?
• Attention architecture - Are critical rules in the right cognitive zones?
• Actionable refactoring - Concrete improvements using YOUR existing examples

Prerequisites

You need: An existing AI configuration file with your project-specific conventions already documented.

You don't need: Tech stack detection, framework templates, or setup wizards.