Ask your codebase anything. See everything that matters, instantly.
Two new features are live in Comper this month. One lets anyone ask your codebase questions in plain language and get answers grounded in actual development history. The other lets you define and surface the metrics that matter to your organization in real time. Both are available now.
Comper Prompt
Now live.
Ask your codebase questions in plain language. Comper uses the full development history as context, so answers are grounded in actual architecture, actual contributors, and actual risk.
Comper Prompt is not a generic AI assistant, nor a search bar. It is a direct line to what your codebase already knows.
Example questions you can ask now:
Where is our highest bus factor risk right now?
Which services have had the most unreviewed changes this month?
What would break if this dependency was removed?
Who are the last three contributors to this module, and when did they last commit?
Answers come back in seconds, not weeks. Saving you hours of unnecessary guesswork and goose chasing.
Custom Statistics
Define the metrics that matter to your organization and surface them in real time.
Standard engineering dashboards are built for engineers. Custom Statistics is built for the conversations that happen above the pull request: board reviews, due diligence, post-incident analysis, and quarterly planning.
What you can configure:
Bus factor concentration scores across any set of repositories
Security exposure rates by team, product line, or acquisition
Ownership coverage across your entire software landscape
Configurable visibility for engineering leaders and executives, not just developers.
Coming Soon
Tech Debt
Identify, prioritize, and track technical debt across your entire software landscape. Not just linting warnings. Structural debt: modules that have outgrown their architecture, dependencies that have become liabilities, services that only one person can safely touch.
Service Catalog
A live, automatically maintained catalog of every service in your software landscape. Ownership, dependencies, health, and history. Built from development history, not manual curation.
Questions or feedback: [email protected]
