Community · E2 · artifact verified
Jump to the video moment that answers you
A Chrome extension for YouTube: ask the video a question in plain text, and Jev's typed judgments locate the moment that answers it - the player seeks straight there instead of you scrubbing.
01 · Role in the system
What Jev does here
The extension slices the video's transcript into segments and asks Jev typed questions per segment - does this part address the user's question, how directly - reading calibrated answers to rank and select the moment, then drives the player's seek to that timestamp. Because the judge returns probabilities rather than generated text, the extension can require confidence before jumping and show why the moment was picked. The manifest and background worker keep it entirely client-side, and the options page carries the user's own API key.
02 · Control boundary
Where Jev sits
Jev as the segment judge behind a video QA jump: transcript segments become typed does-this-answer questions, calibrated confidence gates the seek, and the player navigates to the chosen moment without any text generation.
Code owns the loop, permissions, thresholds, validation, and side effects. Jev owns only the bounded judgments described above.
03 · Known limits
What this evidence does not prove
- Depends on transcript availability; videos without captions are out of scope.
- Fresh project (0 stars); single-browser build.
- User's own API key; per-question call cost on long videos.
04 · Attribution
Public sources
This is a Community record: the project was published by a third-party community author.
- jinukuntlaakhilakumargoud-web ↗Community · github · public · checked 2026-10-07