incoherence

Detect and resolve incoherence between documentation, code, and specifications.

Updated Jan 19, 2026
One-click install
npx skills add https://github.com/loriensleafs/brain --skill incoherence-loriensleafs
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: incoherence
Source: https://github.com/loriensleafs/brain/tree/main/apps/claude-plugin/skills/incoherence
Command: npx skills add https://github.com/loriensleafs/brain --skill incoherence-loriensleafs

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve?

Detect and resolve incoherence between documentation, code, specs, and policy, including a reconciliation phase for applying user-provided resolutions.

Core Features & Use Cases

  • Automated incoherence detection across docs and code, including contradictions, ambiguities, gaps, and policy violations.
  • Reconciliation workflow to apply user resolutions in a controlled, idempotent manner.
  • Parallel exploration & verification using configurable agents to surface evidence and validate findings.
  • Use case: In a large project, run the detection to surface issues, then selectively apply resolutions to align docs and behavior.

Quick Start

Run the detection workflow to produce an incoherence report for your repository, then apply resolutions via the reconciliation phase. Example: python3 scripts/incoherence.py --step-number 1 --total-steps 22 --thoughts "Initial survey".

Frequently Asked Questions about incoherence

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

FAQPage Schema
How do I detect incoherence between documentation and code in a large project?

You can detect incoherence by running a step-based detection workflow that produces an incoherence report. This process surfaces contradictions, ambiguities, gaps, and policy violations across your documentation and codebase.

What is doc-code reconciliation and how does it apply fixes?

Doc-code reconciliation is a controlled, idempotent workflow phase for applying user-provided resolutions. It allows you to selectively align documentation and behavior by applying fixes gathered during the detection phase.

How do I run the incoherence detection script with step-based parameters?

Run the incoherence.py script with step-number and total-steps arguments to enable parallel exploration. Example: python3 scripts/incoherence.py --step-number 1 --total-steps 22 --thoughts "Initial survey".

Can I use parallel agents to verify documentation and specification contradictions?

Yes, the workflow supports parallel exploration and verification using configurable agents. These agents surface evidence and validate findings to support the reconciliation of documentation and specification contradictions.

Does this incoherence detection workflow support policy conformance checking?

Yes, policy conformance checking is supported. The detection workflow automatically identifies policy violations alongside gaps and ambiguities, allowing you to apply resolutions that ensure code and documentation adhere to specified policies.

What are the limitations of automated doc-code reconciliation?

Automated reconciliation requires user-provided resolutions for applying fixes, meaning it does not auto-generate corrections. It relies on the incoherence.py script and operates idempotently within a controlled 22-step workflow process.