Reversa¶
Turn legacy systems into executable specifications for AI agents.
You know that system nobody wants to touch? The one that's been running for 10 years, makes money every day, but nobody really knows what it does on the inside? Reversa was built for it.
What is Reversa?¶
Reversa is a specification reverse-engineering framework. You install it inside a legacy project, activate an AI agent you already use, and it coordinates a team of specialists to analyze the code and generate complete, traceable, ready-to-use specifications for any coding agent.
In other words: Reversa turns undocumented code into operational contracts that an AI agent can understand and use to safely evolve the system.
Quick start¶
In the root of the legacy project:
Then open the project in your favorite AI agent and type:
That's it. Reversa takes the wheel and guides you to the end.
What you'll find here¶
-
Why Reversa exists
The problem it solves and why it matters.
-
Installation
Two minutes and you're ready to go.
-
Analysis pipeline
The 5 phases that turn code into specification.
-
Agents
Meet the team: 14 specialists, each with their own role.
Safety guarantee¶
💾 Back up your project before starting
Although Reversa never modifies your files, AI agents can make mistakes. We strongly recommend:
- Version the project in Git — make sure all files are committed before starting the analysis
- Have the repository on GitHub (or GitLab, Bitbucket) — so you have a safe remote copy
- Make a local copy of the folder — a simple
cp -r my-project my-project-backupprotects against any unexpected event
If something unexpected happens during analysis, you can restore the original state with git restore . or from the backup copy.
Reversa never touches your files
Agents write only to .reversa/ and _reversa_sdd/. No file in your project is modified, deleted, or overwritten. Ever.
No API keys
Reversa does not request, store, or transmit API keys from any service. The intelligence comes from the agent you already use in your environment.