Prompt Intelligence StudioFlagship
Quality-control platform for AI prompts.
More detail
Problem
A prompt's quality is invisible until the model's answer already reflects it — by then a request has been spent, and it isn't obvious which ingredient of a good prompt (a clear goal, context, constraints, structure, an output format, a way to check the result) was actually missing.
Approach
Paste a prompt and score it across seven categories — clarity, context, specificity, constraints, structure, output format and reasoning — with issues flagged and an optimised rewrite offered.
Decision
Local Rules mode scores deterministically in the browser, for free, with nothing sent anywhere. BYOK ('bring your own key') mode adds a model-assisted rewrite using a key the visitor supplies, which never leaves the browser.
Notable detail
The local scorer has no React coupling and no runtime dependencies — a portable pure function, which is what makes it possible to run the same logic in this portfolio's hero.
- 203
- tests passing across the full suite
- 7
- prompt-quality categories scored
- 16
- of those, on the local analyser alone