Hey PH! 👋 Maker here. I built TokenLens after noticing how much of my Claude API spend was pure waste — politeness phrases, redundant context, vague instructions. Most prompts are 30-70% bloat and you can’t see it until you break it down. What makes TokenLens different from other prompt tools: it doesn’t just hand you a shorter prompt. It tells you why each change was made and what principle to apply next time. After a week of using it you stop needing it for simple prompts — because you’ve internalized the patterns. A few things I’d love feedback on: — Does the score card (clarity / conciseness / structure / specificity) feel useful or gimmicky? — Is Challenge Mode something you’d actually use to practice? — What type of prompts would you most want to analyze? Free plan is genuinely useful — 3 analyses per day, no credit card. Jump in and let me know what you think 🙏

TokenLens

See every token. Fix every prompt. Learn why it works.

See every token. Fix every prompt. Learn why it works.