Scores YouTube creators for sponsor-readiness, attaches verified business contacts, detects sponsor history. No charge for records that fail your quality gate. Watchlist mode tracks creator momentum over time.
1.0 (2026-05-14) — Cleaner logs and no missing records under parallel load
Internal parser warnings from the YouTube library no longer flood the run log under parallel processing. Your application logs are no longer truncated.
Every input channel now produces a dataset record, even when an unexpected error fires inside an error handler. No more silently-dropped channels at large cohort sizes.
Failure webhook registration now uses an idempotency key so repeat runs no longer stack duplicate hooks.
1.0 (2026-05-14) — Significantly faster on large channel lists
Channels in a list are now processed in parallel instead of one at a time. A 35-channel run should drop from around 6 minutes to under 90 seconds with default settings.
Transcript fetches inside a single channel also run in parallel now, saving more time on channels with several videos sampled.
New input field lets you tune how many channels run at the same time. Default 4. Drop to 1 if YouTube starts blocking your IP range.
1.0 (2026-05-14) — Fixed empty social links and wrong-channel resolution for handles
Social links from the channel About tab now extract correctly. Twitter, Discord, Instagram and external website links populate per channel.
Channel handles like @mkbhd now resolve to the correct creator. Previous build sometimes routed ambiguous handles to a similarly-named channel.
1.0 (2026-05-14) — Bug fix: sponsor detection was always returning 'never' because three data sources were silently broken
Fixed video transcript fetching. Transcripts were silently failing because the error log was at debug level and the call was using a flow that returned null instead of surfacing the real error.
Fixed video descriptions. Recent video descriptions were always null because the channel-listing only returns short metadata. The fix pulls full descriptions during the transcript sampling pass.
Fixed channel About-page parsing. Business email, social links, and external website extraction now read from the About tab properly. Previous build looked at the wrong path so all three came back empty.
1.0 (2026-05-14) — Bug fix: subscriber count was returning null for most channels
Fixed subscriber count extraction. Previous build returned null for channels using the new YouTube page-header format (which is most of them), causing them to all skip the quality gate.
Quality-gate skips on missing subscriber data now report 'subscribers-unavailable' instead of 'minSubscribers<N' so you can tell a real filter from a data-extraction issue.
