Submit an async Pro deep dive (multi-signal attribution job)

Queue an asynchronous Pro deep dive on ctscout.dev: the full multi-signal attribution run (CT warehouse + DNS, RDAP, homepage, IP/ASN corroboration) executed by a batch worker, via POST /jobs. Returns a job receipt immediately, NOT results.

Asynchronous, Pro only:

• The call returns as soon as the job is queued ({job_id, status: "queued", submitted_at}). Nothing is attributed yet.

• Poll with ctscout_get_job using the returned job_id. Wait about 30 s before the first poll, then back off toward 5 min between polls; the batch worker picks up queued jobs every few minutes and a deep dive can take several minutes to run.

• Requires a Pro API key. A free key gets HTTP 403 with the API's upgrade text (Pro is $49/month from https://ctscout.dev/#tiers). Quota: 20 submissions per key per day (HTTP 429 over). Submitting is not idempotent — a retry queues a second job.

Args:

• company_name (string, optional): organization name, matched exactly as in ctscout_search_company (partial, case-insensitive; 2–200 chars).

• seed_domain (string[], optional): known apex domains to pivot from, max 10. At least one of company_name / seed_domain is required; both may be given. Validated exactly like /scan.

• response_format ('markdown' | 'json', default 'markdown'): a receipt with polling guidance, or the raw 202 body.

Returns (on success, structuredContent follows the declared outputSchema; a failed call — 401, 403, 429, timeout — is isError with no structuredContent): { "job_id": string, // opaque; pass to ctscout_get_job "status": "queued", "submitted_at": string, "poll": "/jobs/" // informational }

What the finished result contains (read it with ctscout_get_job):

• The deep-dive result shape (a /scan never carries it): "domains" of attributed apex domains, each with "attributed_to", an "enrichment" object (confidence_band, weight_total, matched_via, evidence, signal_health, vlm_status, vlm_override) and the underlying discovery evidence under "base"; plus "entity", "run_metadata", "source" and "signals_degraded".

• Plus "snapshot": the warehouse date (YYYY-MM-DD) the deep dive read from. It is present on every deep-dive result because the batch worker sets it, together with "worker_version" and "signals_attempted"; a /scan answer carries its own snapshot from the API.

• "Attributed" means the organization is what the evidence names for that domain (certificate subject, corroborated by the enrichment signals), not an ownership claim. "Candidate" means a semantic name-similarity guess that is NOT an attribution; a deep dive reports attributions with a confidence band, never bare candidates.

• Visual brand verification (VLM) is NOT included in v1: vlm_status stays "pending" or "skipped" and never vetoes a band.

Examples:

• Use when: "Run a full attribution deep dive on CNA Financial" -> { company_name: "CNA Financial" }

• Use when: "Deep-dive from these seed domains" -> { seed_domain: ["cna.com", "cnasurety.com"] }

• Don't use when: you want an answer now — ctscout_search_company / ctscout_lookup_domain are synchronous. Don't resubmit while a job is queued or running; poll it.

Corrections:

• Wrong attribution? Open an issue at https://github.com/minghsuy/ctscout-mcp/issues. Missing entity? Email pro@ctscout.dev. Include the job id, the spec submitted and why the result read with ctscout_get_job is wrong.