reversa-depth-inspection

Inspects a feature across specs, code, tests, and data with parallel analysis lenses to register confirmed bugs.

Updated May 22, 2026
One-click install
npx skills add https://github.com/GleisonOliveira/dockerpilot-mcp --skill reversa-depth-inspection-gleisonoliveira
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: reversa-depth-inspection
Source: https://github.com/GleisonOliveira/dockerpilot-mcp/tree/main/.agents/skills/reversa-depth-inspection
Command: npx skills add https://github.com/GleisonOliveira/dockerpilot-mcp --skill reversa-depth-inspection-gleisonoliveira

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? When a feature keeps breaking, fixing one bug at a time is not enough. This Skill performs a deep, systematic inspection of an entire feature—mapping its specs, code, tests, and data—then sweeps it with specialized lenses to surface every confirmed defect as a registered, traceable bug. ## Core Features & Use Cases - Feature Mapping: Builds a map linking spec sections, implementing code, existing test coverage, and touched data stores before any analysis begins. - Parallel Specialized Lenses: Runs mandatory lenses (spec conformance, data flow, contracts, error states, test coverage, concurrency) plus conditional lenses (security, performance, config drift, observability) as parallel subagents. - Diagnosis-Only Discipline: Never modifies project code; findings are deduplicated, confirmed against observable deviations, and only then registered as bugs in _reversa_bugs/ with full traceability. - Use Case: A checkout feature keeps failing in production. Run this Skill to map the feature, sweep it with all lenses, and receive a consolidated report plus registered bugs for every confirmed defect. ## Quick Start Ask the agent to run a deep inspection on the feature that keeps causing problems, for example: run a depth inspection on the checkout feature.

Frequently Asked Questions about reversa-depth-inspection

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

FAQPage Schema
How do I find all bugs in a feature instead of fixing them one by one?▼

Run a depth inspection that first maps the feature's specs, code, tests, and data, then sweeps it with specialized lenses covering conformance, data flow, contracts, error states, test coverage, and concurrency. Confirmed findings are registered as individual traceable bugs.

What does a spec-to-code conformance inspection check?▼

It compares the implemented behavior against the effective spec (original plus active addenda) to find divergences. Each finding includes evidence such as file and line references, spec excerpts, and a confidence level.

Does this inspection modify or fix my code?▼

No. The skill is diagnosis-only and never edits, deletes, or refactors project files. It writes exclusively to the bug registry directory, creating bug records, inspection reports, and generated views.

When should I use a depth inspection instead of a single bug fix?▼

Use it when a feature repeatedly produces problems and isolated fixes have not helped. It is designed for maintenance-phase investigations where defects may share a structural root cause across specs, code, and data.

How are inspection findings confirmed before becoming bugs?▼

A finding becomes a bug only with an observable deviation between expected and actual behavior, or static proof with a complete causal path. Suspicions and low coverage remain in the report with promoted_to set to null.