Commerce-Intent API

Quickstart

Five-minute integration

Request partner credentials and onboarding, issue one POST per user task, merge the optional extension field, run the call in parallel with your own work.

1. 01

   API key & onboarding

   Email api-actify@outlook.com. You receive onboarding guidance plus ACTIFY_API_KEY and ACTIFY_BASE_URL (production default above).

2. 02

   Make one call

   curl -X POST "$ACTIFY_BASE_URL/v1/match" \
     -H "Authorization: Bearer $ACTIFY_API_KEY" \
     -H "Content-Type: application/json" \
     -d '{"task_query": "which microphone is best for podcasting?"}'

3. 03

   Attach to your response

   Run the Actify API call in parallel with your own LLM or business logic. Merge at the end. Use a finite client timeout tuned to when you send the response — ~1–2 s is a sensible default for many services (shorter caps fail-open more often on fast paths; see API reference).

   import asyncio
   import os
   import httpx
   
   ACTIFY_BASE = os.environ.get("ACTIFY_BASE_URL", "https://actify-api.vercel.app")
   ACTIFY_API_KEY = os.environ["ACTIFY_API_KEY"]
   
   async def call_actify_match(task: str) -> dict | None:
       try:
           async with httpx.AsyncClient(
               timeout=float(os.environ.get("ACTIFY_TIMEOUT_S", "1.5")),
           ) as c:
               r = await c.post(
                   f"{ACTIFY_BASE}/v1/match",
                   headers={"Authorization": f"Bearer {ACTIFY_API_KEY}"},
                   json={"task_query": task},
               )
               return r.json() or None
       except Exception:
           return None  # fail-open: never block your core answer
   
   @app.post("/answer")
   async def answer(req: Req):
       ad_task = asyncio.create_task(call_actify_match(req.query))
       answer  = await your_llm.respond(req.query)
       ad      = await ad_task
       if ad and ad.get("_commerce_layer"):
           answer["_commerce_layer"] = ad["_commerce_layer"]
       return answer