Agent Query API

What this API does

The Agenticfeed passive feed lets AI agents crawl product data by following URLs. The Agent Query API is the active layer. Your AI sends a plain English description of what a buyer needs and gets back the best matching products from a merchant's catalogue — with a reason for each match that you can pass directly to the buyer.

The matching is done by Claude. You supply your own Anthropic API key in the request body, so you pay Anthropic directly for the AI compute. Agenticfeed charges nothing for the AI calls — only for access to the product data.

The response is a schema.org ItemList of Product entries. Every product URL is UTM-tagged so the merchant can see traffic and orders originating from AI agent recommendations in their analytics.

Getting started

1. Sign up for Agenticfeed Create an account at agenticfeed.ai/login.html. Your account gives you access to the product data layer.
2. Get your Anthropic API key Sign up at console.anthropic.com and create an API key. You will pass this in every query request. Agenticfeed uses it to run the matching call — you pay Anthropic directly at their standard rates.
3. Contact us for an API access key Email sales@bluestratus.com to request an Agenticfeed API key. This identifies your application and controls access to the query endpoint. Include a brief description of what you are building.
4. Make your first query Send a POST request to /agent/query with your buyer's intent. See the request format below.

Authentication

Every request requires two keys:

1. Agenticfeed API key

Passed as a Bearer token in the Authorization header. This is your access credential for the Agenticfeed data layer. Contact us to get one.

Authorization: Bearer af_your_api_key_here

2. Anthropic API key

Passed in the request body as anthropic_api_key. This is your own key from console.anthropic.com. Agenticfeed validates it with a minimal one-token call before running the query. If invalid, the request is rejected immediately before any product data is fetched.

You are billed by Anthropic for each matching call at their standard rates for claude-haiku-4-5-20251001. A typical query costs a fraction of a cent.

Query endpoint

POST https://agenticfeed.ai/agent/query

Request body

{
  "query": "ergonomic chair for someone with lower back pain, works from home 10 hours a day, small home office, budget under £300",
  "anthropic_api_key": "sk-ant-your-key",
  "customer_guid": "a9a8378c94",
  "limit": 5
}

Full request example

curl -X POST https://agenticfeed.ai/agent/query \
  -H "Authorization: Bearer af_your_api_key" \
  -H "Content-Type: application/json" \
  -d '{
    "query": "board game for a family with children aged 8 to 12, 3 to 5 players, under £35",
    "anthropic_api_key": "sk-ant-your-key",
    "limit": 3
  }'

Response

Returns application/ld+json — a schema.org ItemList of matched Product entries ordered by relevance. Each product includes a match_reason field explaining why it fits the buyer's need. This is designed to be passed directly to the end buyer by the calling agent.

{
  "@context": "https://schema.org",
  "@type": "ItemList",
  "query": "board game for a family with children aged 8 to 12...",
  "numberOfItems": 2,
  "itemListElement": [
    {
      "@type": "Product",
      "position": 1,
      "name": "Ticket to Ride Europe",
      "url": "https://merchant.com/ticket-to-ride?utm_source=agenticfeed&utm_medium=ai-agent",
      "image": "https://merchant.com/images/ttr.jpg",
      "category": "board games",
      "offers": {
        "@type": "Offer",
        "price": "34.99",
        "priceCurrency": "GBP",
        "availability": "https://schema.org/InStock"
      },
      "agenticfeed": {
        "product_id": "b3f1e29a4c",
        "match_reason": "Suitable for 2 to 5 players aged 8 and up, plays in 45 to 75 minutes, and falls within the stated budget."
      }
    }
  ]
}

If no products match the query, itemListElement is an empty array and numberOfItems is 0. No error is returned.

UTM attribution

Every product URL in the response is automatically tagged with UTM parameters:

utm_source=agenticfeed&utm_medium=ai-agent

When a buyer clicks through and makes a purchase, that order is attributed to agenticfeed in the merchant's analytics. This lets merchants measure exactly how much revenue AI agent recommendations are generating.

Agent Card

Agenticfeed publishes an A2A-compatible Agent Card at a standard well-known URL. AI platforms that support Google's Agent-to-Agent protocol can discover and call the query endpoint automatically using this file.

GET https://agenticfeed.ai/.well-known/agent.json

The card describes the agent's capabilities, accepted inputs, output format, and authentication requirements in a machine-readable format.

Open standard

The feed format this API is built on is published as an open specification. You can read the full standard, implement your own compatible feed, or contribute at:

github.com/bluestratus/agenticfeed

The merchant docs covering the rel="agenticfeed" discovery tag and feed format are at agenticfeed.ai/docs.html.

Questions

Email sales@bluestratus.com to request API access or ask about integration. Include a brief description of what you are building.