Process this YouTube source with the repo-tracker workflow and finish the Airtable catalog update. 

Fetch and review the source title, channel, date, full description, and full transcript. Extract every GitHub repository explicitly listed in the source, plus any distinct launched product that meets the repo-tracker’s criteria. Classify candidates only into the existing authorized tracker tables.

Read the current Airtable schema before writing, including live schema changes not reflected in documentation. Check each candidate’s exact primary URL against the destination table. Exclude exact matches from the insert batch, leave existing records unchanged, and never use fuzzy matching.

For each new row, fill every field supported by source evidence or verified public information. Leave unsupported, unavailable, unverified, or incompatible select fields blank rather than inventing values or creating select options. Verify each destination URL resolves before marking it verified. Populate agent-facing fields only when the source and relevant local inspection support them. Never write to read-only URL button fields, create tables, rename tables, or restructure the base.

After inserting, read each created record back and verify its key fields. Report all extracted candidates, rows inserted, exact-URL duplicates, fields left blank and why, and candidates that do not fit the approved tables. Do not claim completion if any item or verification remains unresolved.