# Lyriks Community Edition
**Turn code into specs. Find gaps. Understand what a change affects.**
[](https://github.com/lyriks-io/lyriks-community/actions/workflows/ci.yml)
[](https://github.com/lyriks-io/lyriks-community/actions/workflows/publish-image.yml)
The Control Center, open from any page: the scores, the checks behind them, and the next thing to fix.
Lyriks is a self-hosted workspace that connects a product's features, rules,
screens and data in one specification. Your AI coding client reads the code
and writes the spec through MCP (Model Context Protocol). You review it in
the browser and use Lyriks to check for gaps and contradictions.
## What you can do with it
Each of these works from the spec and evidence recorded so far. None of them
replaces tests of the running application.
### Code to spec
You need to know what a product does, and only its code can tell you.
Lyriks turns what your AI client reads in that code into a spec, with links
back to the source.
### Coherence
You need to know whether features, permissions and rules contradict each
other or leave gaps.
Lyriks checks the recorded spec and reports the issues to review and fix.
### Impact analysis
You need to know what to review before changing a feature or rule.
Lyriks exposes the links between features, screens, rules and data, so you
and your AI client can see what the change may affect.
### Implementation analysis
You need to know how much of the spec is reflected in the code.
Lyriks uses the code evidence supplied by your AI client to report mapped
elements, missing implementation and changes since the last sync.
## What leaves your machine
Lyriks stores your projects locally and makes no automatic calls to public
services by default. Activation uses an email address supplied on