Finds contacts that a completed Full Pipeline scan job saved to a BDR's prospect_contacts but that never got tagged into the hypothesis group the scan was run for (the run stopped or errored between the scan and the tagging step). They're already paid for. Two ways to handle them: run an automatic sweep (same Target Contact / Exclude Rules classifier the pipeline uses, tags the keepers, archives the rest), or browse & curate manually — search, export, paste a list of URLs you already know you want, or run the AI classifier just for suggestions and decide row by row yourself. Rule-dropped contacts always land in Rule-Dropped Contacts Report either way.
Run the dry run first — it tells you how many untagged contacts exist per group and BDR. Then untick "Dry run" to apply. In a dry run the rules are not applied (no AI cost), so the "would be rescued" number is before the rules filter. Safe to re-run: it only ever touches contacts that are still untagged. Want to review individual contacts before deciding, instead of an automatic sweep? Use the Browse & Curate tab above.
Same candidate pool as the Automatic Sweep tab (untagged, not excluded-from-campaign, from this group's completed scan jobs) — but nothing is written until you say so. Already-connected contacts are hidden by default (net-new only) — check the box to see them anyway (they're never auto-kept).
Paste one LinkedIn URL per line, or a comma-separated list. Any loaded candidate whose URL matches gets marked Keep. URLs that don't match any loaded candidate are reported so you can double check.
Runs the same classifier the pipeline uses (~1 AI call per 50 contacts) against every currently loaded candidate. Defaults to this group's saved rules — edit below for an exploratory pass without changing the group's saved rules. Shows suggestions per row; your manual Keep marks are untouched unless you choose to bulk-apply them afterward.