How does ELD-debug improve software quality?

By prioritizing 'quality over speed' and requiring logical proof of a fix, it prevents shallow 'band-aid' solutions and ensures that the underlying architectural laws are restored and documented.

When should I use the eld-debug skill?

Use it when facing complex, non-obvious bugs, system regressions, or when you need a highly documented and verified debugging process for mission-critical code.

What is the ELD debugging methodology?

ELD debugging treats software bugs as 'deviations from laws.' It uses a systematic loop to observe symptoms, model the broken law as a logical expression, and verify the fix using multiple levels of evidence.

What is the 'Evidence Ladder' in this skill?

The Evidence Ladder is a framework for verifying fixes across five levels: L0 (Static Analysis), L1 (Unit Tests), L2 (Integration Tests), L3 (Fuzzing/Chaos Testing), and L4 (Production Telemetry).

ELD Law-Driven Debugging

Name: ELD Law-Driven Debugging
Author: CAPHTECH

byCAPHTECH

0•

セキュリティとテスト

Systematically resolves software bugs by treating them as deviations from logical laws and verifying fixes through an evidence-based loop.

ELD Law-Driven Debugging is a specialized Claude Code skill that transforms the debugging process from reactive troubleshooting into a rigorous, scientific methodology. By viewing bugs as violations of system 'laws' or invariants, it guides developers through a structured six-phase loop: Sense, Model, Predict, Change, Ground, and Record. This skill is ideal for complex root cause analysis where evidence-based verification is required, ensuring that fixes are not just functional but logically sound and verified against a multi-level Evidence Ladder.

主な機能

01Evidence Ladder verification from static analysis (L0) to production telemetry (L4)

02Pattern-based recording for long-term knowledge accumulation

030 GitHub stars

04Six-phase systematic debugging workflow (Sense to Record)

05Logic-based law formulation to define expected system behavior

06Impact prediction analysis to minimize side effects and regressions

ユースケース

01Performing root cause analysis on complex concurrency or state-management issues

02Investigating intermittent production failures using structured telemetry and logs

03Implementing high-integrity fixes for critical systems that require formal verification

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add caphtech/claude-marketplace eld-debug

For use in Claude.ai and ChatGPT

主な機能

01Evidence Ladder verification from static analysis (L0) to production telemetry (L4)

02Pattern-based recording for long-term knowledge accumulation

030 GitHub stars

04Six-phase systematic debugging workflow (Sense to Record)

05Logic-based law formulation to define expected system behavior

06Impact prediction analysis to minimize side effects and regressions

ユースケース

01Performing root cause analysis on complex concurrency or state-management issues

02Investigating intermittent production failures using structured telemetry and logs

03Implementing high-integrity fixes for critical systems that require formal verification

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add caphtech/claude-marketplace eld-debug

For use in Claude.ai and ChatGPT