Step back before you burn out.
Four brief check-in questions help surface mental strain. Alongside your work patterns, they inform a recommendation to continue, clarify your goal, take a break, or stop.
An AI-powered VS Code extension
Avoid burnout.
Waste fewer
AI tokens.
An extension that helps you manage your mental state while working with AI.
Get a clear next step: continue, clarify your goal, take a break, or stop. See your token use and review unread AI changes in one shareable work summary.
Free research preview · Local tracking works without cloud participation
A nudge before the next run.
Interactive exampleVisual Studio Code
AI Work-State Check-in — 1 of 4
Working with AI left me mentally drained.
- Never
- Rarely
- Sometimes
- Often
- Almost always
Four brief questions about mental drain, goal clarity, review understanding and reactivity.
Therapeutical.ai: Your next step
A good moment to call it.
84% estimated risk of unproductive AI work or worsening work state over the next two hours. Experimental.
Stop for today · Open summary
Your first seven active days build a baseline. Follow-up check-ins help refine your estimates.
Therapeutical.aiExample session
How it helps
Break the cycle of wasted tokens.
See token use alongside repeated attempts and unread changes. Use your next-step recommendation to clarify the task or pause before another run.
Install for VS Code
Free research preview · No user API key
Why it matters
Your mental state affects how much AI work costs.
When you feel drained, it can be harder to judge changes or decide what to try next. Therapeutical.ai measures you + AI together to help you spot when effort stops paying off.
Private by design. No source code, prompts, conversation text or file names in research data.
FAQ
Questions about measurement, privacy and what the recommendation means.
What does Therapeutical.ai measure?
Token use, repeated attempts, discarded work and review activity—paired with four brief AI Work-State questions.
Why measure mental strain and tokens together?
Work patterns show repeated attempts and review activity. Your check-in adds how you feel, helping you choose whether to continue, clarify, break, or stop.
Does it diagnose burnout?
No. It is an experimental work-awareness tool, not a medical device or diagnostic service.
Does it read my code or prompts?
No. Code, prompts, conversation text, file names and paths are excluded from research and prediction data.
When do predictions start?
With cloud participation enabled, suggestions start after your first check-in. Check-ins and predictions continue about every two hours while you work, including after the first week. Model percentages start once enough completed follow-ups include both smoother and more difficult work periods. Early estimates are experimental.
Is it free?
Yes. The current research preview is free, and local tracking works without cloud participation.
Make your next AI session count.
Protect your focus. Spend tokens when the work is worth it.
Install Therapeutical.ai