Glossary

Campaign Studio concepts — the “why” reference.

Campaign

A (client, mode) unit of discovery + curation work. Every target lives inside exactly one campaign. Mode is fixed at creation: prompt, keyword, or persona.

Target

A discovered prompt, keyword, or persona. Type is implicit from the parent campaign's mode. Status flows wishlist → accepted → rejected. Accepted targets are the primary contract for downstream tools.

Classification

A per-campaign cluster label grouping related targets. Either auto-generated (clustering-driven) or user-provided (future). Sprint 3+.

Expand / Refresh / Rebuild / Archive

The four campaign-level actions. Expand adds new targets only. Refresh updates existing targets. Rebuild wipes and re-runs discovery (snapshot preserved in campaign_runs). Archive soft-hides a campaign from list views. Sprint 2+.

audit_site_id

Opt-in link from a campaign to an ai-content-audit-tool `sites.id`. When set, discovery actions consume the audited site's crawl + embeddings + ideas check instead of re-crawling. MVP requires this for any mode needing page data.

Hub client (hub_client_id)

The client identity propagated from hub.hivedigital.com. Primary scope key — every campaign row carries it; every query filters by it. NULL when the tool runs standalone (dev/testing).

Sprint 1 stub. Full glossary ships alongside mode / action UX in Sprint 2+.