slopcheck-jev
A prose linter that catches AI writing tells. Regex settles the 18 a pattern can settle. Jev takes the 15 that need reading, as 15 Nouls in one call, 604 ms median. It ships as a Claude Code
Stop
hook that scores Claude's own output after every turn and warns rather than blocks.
Where Jev makes the decision
Deleting a feature made it more accurate. The first build made a second Jev call to pin each tell to a sentence. Removing it took precision from 0.80 to 0.95 and latency from 1,365 ms to 575 ms. Jev returns typed answers rather than text, so a quote has to come from a second round of typed questions about lines, and that round was where every defect lived. The README documents the two intermediate designs that were worse.
Why it matters
Adds structured choices or scores to the workflow; performance and cost benefits have not been independently verified.
Claims & evidence
Author and repository text is preserved where clear; missing languages are enriched automatically. Checks establish source-level Jev integration, not runtime, safety, or performance validation.
About Jev
Jev is TypeSafe AI's System One decision model: it returns typed scores and choices (not free text), so agents can route, filter and gate at low latency and cost.
Source: harshpuri84/slopcheck-jev — 0 stars, MIT. Listed by Awesome Jev.
