ACR Draft
Write the accessibility report (VPAT) a university asks for before it buys your software: scans plus a guided review, exported to Word.
Apps / Prompt Shrink Ray
Compress an LLM prompt with a live token estimate and a check that the keywords survived.
Trims long AI prompts so they cost less to send. Paste a prompt and it splits it into parts (instructions, background, examples, the task) and tightens each at the level you choose: removing filler and hedging words, shortening wordy phrases, dropping instructions you repeated and cutting extra examples. Code blocks are left exactly as written. You see a word-by-word comparison, the estimated token and cost savings, and a list of any numbers, names or quoted text a change removed, so you can check nothing important went missing.
Token counts are estimates (about four characters per token), not exact counts.
Write the accessibility report (VPAT) a university asks for before it buys your software: scans plus a guided review, exported to Word.
Paste a random sampling function and see exactly how biased it is.
Measure how much code a change adds in syntax-level units instead of lines.