bidirectional-differential

Run parallel Claude sessions to classify drift between EARS and code.

90|14|Updated Apr 11, 2026

Search Tags:#Claude #code #coherence #EARS #arrow-maintenance #differential-audit

Ready to run

npx skills add https://github.com/jszmajda/lid --skill bidirectional-differential

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: bidirectional-differential
Source: https://github.com/jszmajda/lid/tree/main/plugins/lid-experimental/skills/bidirectional-differential
Command: npx skills add https://github.com/jszmajda/lid --skill bidirectional-differential

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Audits coherence between EARS and code by running two parallel Claude sessions — one reconstructs code from an EARS, the other reconstructs the EARS from the code — then classifies drift to surface intent gaps and mismatches.

Core Features & Use Cases

Surface alignment gaps between specifications and implementation by parallel session runs.
Map user-described scopes to arrow overlays, EARS IDs, and segment metadata to drive targeted audits.
Produce per-EARS audit records in docs/arrows/_experiments/bidirectional-differential for traceability and review.

Quick Start

Invoke the /differential-audit command with one or more EARS IDs to start a bidirectional-differential audit.

Frequently Asked Questions about bidirectional-differential

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

FAQPage Schema

How do I audit drift between EARS specifications and source code?▼

You can audit EARS and code drift by running two parallel Claude sessions that reconstruct specifications from code and code from specifications, then classify mismatches to surface alignment gaps.

What is the bidirectional differential method for checking specification coherence?▼

The bidirectional differential method checks specification coherence by spawning A- and B-direction sessions to reconstruct code from EARS and vice versa, classifying drift to resolve inputs and surface intent mismatches.

How do I run a differential audit using Claude for a specific EARS ID?▼

To run a differential audit, invoke the /differential-audit command with one or more EARS IDs, which triggers a six-step protocol to resolve scopes, spawn parallel sessions, classify drift, and write audit artifacts.

Can I scope a differential audit to a specific arrow overlay and segment metadata?▼

Yes, you can scope audits to a specific arrow and EARS by mapping user-described scopes to arrow overlays, EARS IDs, and segment metadata to drive targeted parallel session runs.

Where does the bidirectional differential audit store per-EARS traceability records?▼

The bidirectional differential audit stores per-EARS audit records in the docs/arrows/_experiments/bidirectional-differential directory for traceability and review.

What are the limitations of using parallel sessions to detect EARS and code mismatches?▼

A limitation of using parallel sessions is that the audit requires resolving inputs and stripping identifiers before spawning sessions, meaning unresolved or poorly scoped arrow overlays may affect drift classification accuracy.