1.0Score
A score for every session
ROIP rates how well you prompt across a whole conversation, not just one message, on a scale of 0 to 100. Follow it week to week and see which habits move it.
See how effectivelyyou work with ChatGPT, Claude, Gemini, Copilot, DeepSeek, Grok, Mistral AI, Kimi, Muse, Llama, Qwen, TypeSafe AI, Clef, and SubQ
Understand your AI use. Improve how you prompt.
Built for the frontier models your team already uses.
Every prompt is an experiment. Worsley keeps the results.
Your best prompts, and your worst ones, disappear when you close the chat. Worsley keeps a record of every session, in your browser and in your terminal, and scores it, so you can see what works and what doesn’t.
2.0Usage
Where your time goes
Time, sessions and turns, split by model and by project. Find out whether you really use Claude more than ChatGPT, or just think you do.
3.0Sessions
Every conversation, turn by turn
Open any session and see each prompt scored, where the conversation drifted, what worked, and what to try next time.
4.0Teams
For teams
See how your team works with AI across people and projects, and what to coach next.
- Team analytics
- Usage, ROIP and trends across people and projects.
- Project view
- Where your team's time goes in shared ChatGPT projects. Members see scores only for their own sessions.
- Team insights
- Which criteria the team is strong and weak at, and what to work on next.
- Access controls
- Team settings decide who can invite and whether members see ROIP.
Typing speed was once a line on a résumé. AI skill will be next, and we’re defining how it’s measured.
See how your next AI session scores.
Install the extension and connect your terminal, then keep working in ChatGPT, Claude, Gemini, Copilot or your terminal. Worsley scores each session and tells you what to change next time.
Shared calendar launch announcement
5 turns, scored today