Prompts, managed like software
Everything you need to author, review, test, and ship LLM prompts with confidence.
Version control for prompts
Every edit creates a new version with an author and a note. Browse the full timeline, view any past version, and roll back to a known-good one in a single click. Nothing is ever overwritten.
Diffs that show impact
Compare any two versions with line- and word-level diffs in split or inline view. The change is shown next to its eval-score delta, so you see not just what changed but whether it helped.
Testing & evaluation
Write test cases with friendly checks — must-mention, valid JSON, no personal data, length limits, or an AI judge. Run them in the dashboard or in CI, and get a pass-rate gate that blocks regressions.
Edge delivery & environments
Promote versions across development, staging, and production. Your app fetches the live prompt from a fast, cached runtime API — so you update prompts without shipping a new build.
API, CLI & GitHub Action
A clean REST API, the pvcs CLI for your terminal, and a GitHub Action that runs your eval suite on every pull request and fails the check when quality drops.
Built for teams
Invite teammates with roles, share a workspace, and track every change in the activity log. Workspace-isolated data and per-key usage keep things safe as you scale.