signalproof-full-debug

Coordinates governed multi-lane debugging for recurring cross-layer defects with evidence and rollback controls.

Updated Aug 15, 2026
One-click install
npx skills add https://github.com/docreo/Signalproof-Skills --skill signalproof-full-debug-docreo
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: signalproof-full-debug
Source: https://github.com/docreo/Signalproof-Skills/tree/main/skills/signalproof-full-debug
Command: npx skills add https://github.com/docreo/Signalproof-Skills --skill signalproof-full-debug-docreo

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Complex defects that recur across chats, span multiple layers, or threaten protected state often fragment into duplicate retries and unsupported fixes. This Skill orchestrates the highest-governance Signalproof debug workflow so failure memory, authorization, role separation, and verification stay intact. ## Core Features & Use Cases - Bounded Full Debug Envelope: Defines workstream, Git basis, symptom, protected state, allowed surfaces, acceptance evidence, recovery path, and STOP conditions before any consequential correction. - Failure Memory Restoration: Searches known-error registries and restores prior failed-attempt counts so retries never reset across chats or cosmetic changes. - Builder/Design/Governor Lane Separation: Runs a spawn protocol with independent or logical lanes that challenge implementation, design drift, scope, and authority. - Exact-Final Artifact Validation: Re-validates the final bytes after every change and retests the original failing condition with regression and protected-state checks. - Use Case: A harness failure has recurred across three sessions and a fix could touch protected main-branch state. Invoke full-debug to restore the failure fingerprint, authorize a bounded envelope, run lane-separated correction, and stop after three materially unsuccessful attempts. ## Quick Start Ask the AI to run signalproof full-debug on the recurring test failure, restoring prior known-error memory and authorizing only the bounded debug envelope before correcting the smallest supported cause.

Frequently Asked Questions about signalproof-full-debug

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

FAQPage Schema
How do I debug a defect that keeps recurring across sessions?▼

Use full-debug to restore the failure class and prior attempt count from known-error memory before retrying. It enforces a bounded envelope and stops after three materially unsuccessful attempts against the same unresolved failure class.

When should I use full-debug instead of ordinary debugging?▼

Use full-debug when defects recur across attempts, span multiple layers, threaten protected state, or need independent challenge from multiple disciplines. For narrow localized defects, ordinary signalproof-debug remains the appropriate choice.

What does authorize all mean inside a full-debug session?▼

It means all owner-authorizable actions within the current bounded debug envelope, subject to protected-state and security rules. It excludes releases, production deployment, protected-main mutation, secrets use, and privilege escalation, which require separate authority.

How does the Builder Design Governor lane separation work?▼

Full Debug runs a spawn protocol with three lanes: Builder implements bounded corrections, Design challenges architecture and drift, and Governor challenges scope, authority, and evidence. Lanes can be independent agents or explicitly labeled logical lanes in one session.

Why does full-debug stop instead of continuing to retry fixes?▼

It stops when retries become cosmetic, protected state is unknown, validation fails, or three materially unsuccessful attempts against the same failure class are reached. This prevents unproductive loops and preserves evidence for an owner decision.