legacy-rescue

What problem does it solve?

Legacy-rescue provides a disciplined, low-risk workflow for making real progress on legacy codebases that are fragile, undocumented, or lacking tests without breaking production. It prevents paralysis-by-fear by establishing a codemap, building characterization tests, planning rollbacks, and enforcing small, reversible changes so teams can safely modernize or add features.

Core Features & Use Cases

• Codemap-first orientation: Generate a module inventory, dependency graph, entry points, and data-flow documents to know what exists before changing it.
• Risk-aware prioritization: Identify feedback-center modules and rank areas from lowest to highest blast radius so changes begin in safe places.
• Safety net and rollback: Produce characterization tests, explicit rollback plans, and strategies such as feature flags or dual-write for stateful migrations.
• Atomic commits and debugging loop: Enforce one-concern-per-commit, integrate code review and monitoring, and iterate through a structured debugging loop when issues surface.
• Use cases: Inheriting an untested production system, modernizing fragile services, adding features to code no one dares to touch, or ghost-fixing critical bugs.

Quick Start

Ask the assistant to run a legacy rescue by generating a codemap, identifying feedback-center modules, creating characterization tests, drafting rollback plans, and proposing a sequence of small atomic commits for the first change.