Documentation

Agent-readable operations for AI visibility teams.

OSPEA monitors how answer engines describe a brand, records the evidence behind each observation, and turns missing source truth into reports and fix briefs that humans or agents can act on.

JSON manifestAPI reference

Live entry points

Agents should start at /agents, then use the manifest and policy files for capability, permission, and trust context.

https://ospea.io/agentshttps://ospea.io/agents/manifest.jsonhttps://ospea.io/agents/capabilities.jsonhttps://ospea.io/agents/pricing.jsonhttps://ospea.io/agents/policies.jsonhttps://ospea.io/agents/tools.jsonhttps://ospea.io/.well-known/ospea-agent.jsonhttps://ospea.io/docshttps://ospea.io/api-reference

1. Define the workspace boundary

Use a workspace as the tenant boundary for brands, prompts, competitors, scan runs, first-party imports, reports, costs, and audit receipts.

2. Seed exact prompts and competitors

SavedPrompt rows preserve the exact buyer prompts OSPEA should repeat. Competitor entities define who is compared in share-of-voice and citation reports.

3. Run provider observations

Visibility snapshots store provider, platform, model version where available, timestamp, full answer text, citations, source classes, latency, cost, and scan status.

4. Read reports before triggering mutations

Agents should inspect share-of-voice, weekly AIO, and scan status endpoints before queuing new scans or importing first-party data.

Observation contract

Fields every AI visibility run should preserve.

provider/platformexact promptmodel/versiontimestampfull answer textcitation URLsnormalized citation domainsource classbrand mentioned/citedmention positioncompetitors mentioned/citedprompt clusterrun typecost and latencyscan status

Reports agents can read first

GET https://api.ospea.io/api/v1/workspaces/{workspace_id}/monitoring/share-of-voice

Weekly entity share-of-voice by platform, prompt cluster, source class, citation class, success count, failure count, mentions, citations, and visibility score.

GET https://api.ospea.io/api/v1/workspaces/{workspace_id}/monitoring/weekly-aio-report

Unified weekly AIO report joining OSPEA AI scans, Google Search Console, and Bing AI Performance imports.

GET https://api.ospea.io/api/v1/workspaces/{workspace_id}/monitoring/weekly-aio-report/export

Downloadable weekly AIO report export.

GET/POST/PATCH/DELETE https://api.ospea.io/api/v1/workspaces/{workspace_id}/monitoring/library

SavedPrompt rows used by repeatable OSPEA AI visibility scans.

GET/POST/PATCH/DELETE https://api.ospea.io/api/v1/workspaces/{workspace_id}/monitoring/competitors

Tracked competitor entities used in AI visibility scans and share-of-voice reporting.

Guardrails for trustworthy automation

  • Do not treat provider timeouts as ordinary zero-visibility answers.
  • Do not fabricate Google Search Console or Bing AI Performance metrics.
  • Do not trigger budget-consuming scans without user or workspace-owner approval.
  • Do not scrape private UI screens when an official report endpoint exists.