A score for every prompt
Clarity of the goal, context, constraints and format, efficiency and risk: each scored, with the reason and a concrete fix.
Early access
Better prompts. Fewer tokens.
Promptlift scores your AI prompts out of 100, tells you what is missing and suggests a better version, from the browser, your terminal or the API.
A free quota that refreshes every four hours. No card needed to start.
A better version, on request
Promptlift rewrites the prompt with what was missing, so you can compare the answers you get.
Your sessions and folders
The promptlift CLI reads a folder or your Claude Code sessions, shows the token cost first and masks secrets before anything leaves your machine.
An AI style check for text
Counts the stock phrases and invisible characters that make text read as machine-written. An estimate of style, not proof of origin.