◇ API & MCP — live

Orbit, for machines
you already trust.

Everything the app does, your agents and automations can do too. Run discoveries from Claude, pull your crew into a CRM, trigger missions from your own stack — same brief, same scoring, no dashboard required.

REST API

Call it from anything

A clean, credit-metered REST API over everything: launch discoveries, list and reveal people, fetch drafts, export crew. Bearer-key auth, same permissions as the app.

// curl, SDKs, your backend
MCP SERVER

Let your agent fly missions

Connect Claude, Claude Code or any MCP-capable LLM to Orbit. Your agent gets typed tools — run_discovery, reveal_creator, export_crew — and initiates work on its own.

// Claude, Claude Code, any MCP client
AUTOMATIONS

Put missions on a schedule

Webhooks for every run event plus native triggers for Zapier, Make and n8n. New hot-iron breakout → Slack ping → contact lands in your CRM, hands-free.

// Zapier, Make, n8n, webhooks
// the docs

The reference — click through it

POST/v1/searches· queries in, run out

Launch a search: hand Orbit a brief and explicit queries — each with its own platforms, and an optional angle and audience. Returns a search id, its run id, the estimated credits, and your plan's reveal budget. Poll the run or register a webhook for run.completed.

# launch a search
curl -X POST https://api.orbitcrew.app/v1/searches \
-H "Authorization: Bearer orb_sk_…" \
-d '{ "brand_id": "brnd_meridian",
"brief": "gear reviewers for our pour-over",
"queries": [{ "text": "pour-over coffee reviews",
"platforms": ["instagram","tiktok","youtube"],
"angle": "ride the breakout",
"audience": "home baristas" }] }'
# response
{ "id": "srch_88", "run_id": "run_47",
"estimated_credits": 12, "reveal_budget": 60 }
POST/v1/searchesLaunch a search (queries in)
GET/v1/searchesSearch history
GET/v1/searches/{id}Per-query analysis
GET/v1/discoveries/{id}Run status + results
// model context protocol

Give your agent
a seat at Mission Control.

One MCP server, every Orbit capability. Ask Claude to run a discovery while you're in a meeting, have Claude Code wire your crew into your CRM, or let an automation platform reveal and export on a schedule — all governed by the same credit balance and permissions as the app.

run_discoverylaunch a search (marketed alias of run_search)
list_creatorsquery surfaced people by fit, signal, platform
reveal_creatorunlock profile, rationale and draft
draft_messageregenerate a draft with a new angle
export_crewsend contacts to CSV, CRM or webhook
get_radar_digestwhat changed in your orbit this week
CLAUDE — CONNECTED TO ORBIT MCPLIVE
Find me 20 creators talking about pour-over problems this week, reveal anyone above fit 85, and add them to the HubSpot list "Q3 partners".
→ orbit.run_discovery · orbit.reveal_creator ×6 · orbit.export_crewDone — run #48 surfaced 23 people, 6 scored above 85. All six revealed with drafts and exported to HubSpot. Total spend: 22 credits. Want the drafts summarised?

Your crew, wherever you work.

Export contacts and drafts to CSV, HubSpot, Notion, Google Sheets — or anywhere via webhooks. The API and MCP server are live today; one-click exports land next.

CSVHUBSPOTNOTIONSHEETSWEBHOOKS
Get your API key →