Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.arcmira.com/llms.txt

Use this file to discover all available pages before exploring further.

​
2026-05 — Commercial intelligence (v1.1.0)

  • New Recommendations API: GET /v1/recommendations and GET /v1/entities/{id}/recommendations. Filter by mention_class (ad_read / endorsement / mention / all), channel_id, channel_name, min_confidence, date_from / date_to, and include_disputed. Public IDs use the com_* scheme.
  • New Channel sponsors API: GET /v1/channels/{channel_id}/sponsors with min_ad_reads (default 3) and status filters. Built on the recurring-sponsor and singleton-cleanup rules that power Arcmira’s internal advertiser dashboards.
  • New Entity discovery: GET /v1/entities/search — fuzzy, ranked, free. Optional has_recommendations_data filter. Distinct from /v1/entities/lookup, which still does exact canonical resolution.
  • GET /v1/mentions?details=full attaches matching commercial recommendations to each appearance row (recommendations.items[] with com_* IDs, verbatim quotes, promo codes, offers, confidence).
  • recommendations_summary block on GET /v1/entities/{id} for organization and product entities — totals, unique_shows, first/last seen, channels_as_sponsor.
  • recommendations_summary teaser on GET /v1/channels/{slug}: sponsor_count for everyone, top_sponsors only for Pro+ callers with recommendations:read.
  • New POST /v1/feedback: submit corrections against the exact query you ran. type=recommendations and type=channel_sponsors auto-apply and re-aggregate brand_profiles; other types are logged for review. Reason codes: false_positive_ad_read, false_positive_endorsement, missed_ad_read, missed_endorsement, wrong_classification, wrong_entity, other.
  • New API key scopes: recommendations:read, recommendations:write. Selectable in Settings → API Keys on Pro+ plans only.
  • New /v1/me field: recommendations_api_enabled (true only on pro_plus, ultra, teams, enterprise).
  • New error codes: recommendations_not_enabled, invalid_feedback_request, channel_not_found, recommendation_not_found. Existing insufficient_scope message now references the new scopes when applicable.
  • Commercial rows (/v1/recommendations, /v1/channels/{id}/sponsors, details=full enrichment items) bill at a higher per-row rate than appearance rows. Feedback and entity search remain free.

​
2026-05 — Entity & mentions surface (v1.0.1)

  • New /v1/entities/lookup: resolve a name (and optional type) to a canonical ent_<id>, following merge chains. Returns merged_from_id when the input was an alias.
  • New /v1/entities/{id}: read canonical entity metadata directly.
  • New /v1/entities/{id}/mentions and /v1/mentions: cross-type mention queries with entity_id / entity_name / entity_type / channel_* / sentiment / is_appearance / date_* / q filters.
  • Non-person appearance routes (/v1/organizations/{slug}/appearances, /v1/products/.../appearances, /v1/topics/.../appearances, /v1/channels/.../appearances) return 400 appearances_person_only. The OpenAPI document only documents /v1/people/{slug}/appearances. See Appearances vs. mentions.
  • Mentions endpoints return 400 appearances_person_only if is_appearance=true is set on a non-person entity.
  • X-Request-Id, X-Arcmira-Version, and RateLimit-* headers stable across all v1 responses.

​
2026-04 — v1.0.0

Initial public API surface:
  • Search and entity reads.
  • Person appearances and per-entity related lists.
  • Watchlists and trackers CRUD with idempotency.
  • /v1/watchlists/{id}/alerts and /v1/trackers/{id}/alerts recent delivery history.
  • Cursor pagination with data / has_more / next_cursor.
  • Stable error envelope and X-Request-Id correlation header.
  • Scoped API keys: read, watchlists:write, trackers:write.