Verified:GitHub

reversa-coding

Transform staged forward execution plans into implemented code with regression guards.

Updated May 5, 2026

Search Tags:#code generation #regression testing #reverse engineering #audit trails #pipeline orchestration #dependency handling #plan to code

Ready to run

npx skills add https://github.com/AlexandrePontesjr/calculadora-Juridica --skill reversa-coding-alexandrepontesjr

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: reversa-coding
Source: https://github.com/AlexandrePontesjr/calculadora-Juridica/tree/main/.agents/skills/reversa-coding
Command: npx skills add https://github.com/AlexandrePontesjr/calculadora-Juridica --skill reversa-coding-alexandrepontesjr

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill converts a staged forward execution plan into real, implemented code while preserving legacy-to-code traceability and setting up regression guards for future runs.

Core Features & Use Cases

Plan-to-code execution by phases: Executes action blocks phase-by-phase (Preparação, Testes, Núcleo, Integração, Polimento) while respecting dependencies and parallelizable subsets.
Strict safety gate via reverse extraction: Requires _reversa_sdd/ to exist and contain architecture.md and domain.md, otherwise it aborts without creating artifacts or touching files.
Audit trails for future confidence: Generates feature-dir/legacy-impact.md and feature-dir/regression-watch.md and appends execution outcomes to feature-dir/progress.jsonl for traceability.
Post-execution integrity hooks: Applies standard before-coding and after-coding hooks as defined by the reversa framework.

Quick Start

Run /reversa-coding when you have already executed /reversa (complete or at least essencial) so that _reversa_sdd/architecture.md and _reversa_sdd/domain.md exist.

Frequently Asked Questions about reversa-coding

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

FAQPage Schema

How do I turn a staged execution plan into implemented code?▼

To turn a staged execution plan into implemented code, a transformation process reads action blocks from actions.md and implements them phase-by-phase while respecting dependencies and parallelizable subsets.

What is legacy-to-code traceability and how does it work during feature development?▼

Legacy-to-code traceability preserves the link between forward execution plans and implemented code by generating audit trails like legacy-impact.md and appending execution outcomes to progress.jsonl for future reference.

Do I need reverse extraction artifacts before generating code from an execution plan?▼

Yes, generating code from an execution plan requires reverse extraction artifacts. The _reversa_sdd/ directory must exist and contain architecture.md and domain.md, otherwise the process aborts without touching files.

How do I set up regression guards and audit trails for staged feature development?▼

To set up regression guards and audit trails for staged feature development, generate regression-watch.md and legacy-impact.md based on architecture and domain rules, while updating actions.md and appending line-by-line to progress.jsonl.

Why does my plan-to-code execution abort without creating any files?▼

Plan-to-code execution aborts without creating files when strict safety gate verification fails, specifically when the _reversa_sdd/ directory or its required architecture.md and domain.md artifacts are missing.

Can I execute parallel action subsets during phase-by-phase code generation?▼

Yes, you can execute parallel sub-sets during phase-by-phase code generation. The process handles dependencies across phases like Preparação, Testes, Núcleo, Integração, and Polimento, allowing optional parallelizable subsets.