Retrieve a specific documentation page by ID from 10 pre-indexed platforms via x402 micropayment.
What it does
AgentUtil Docs `/lookup` is a POST endpoint that retrieves the full chunked content of a specific documentation page by its identifier and platform. It is designed as a follow-up to the `/query` semantic search endpoint: once an agent identifies a relevant page via search, `/lookup` returns all chunks for that page with source URLs and section labels. The service covers approximately 38,000 chunks across 10 pre-indexed platforms including Shopify, Stripe, Cloudflare Workers, Next.js, Anthropic API, OpenAI API, Tailwind CSS, htmx, x402, and Jesse.
The endpoint accepts a JSON body with required `platform` and `page` fields, plus an optional `type` filter (e.g. mutation, query, function). It returns an array of chunk objects each containing content text, source URL, and section name, along with a count of chunks returned. Pricing is $0.002 per request paid via the x402 protocol using USDC on Base. A free tier of 10 requests per day per IP is available with no signup required.
The probe returned HTTP 400 (not 402) on POST, which suggests the endpoint expects a valid JSON body to trigger the x402 payment challenge rather than responding with 402 on an empty request. The OpenAPI spec and landing page are well-structured, and the service exposes standard agent discovery files (agent-service.json, agent.json, openapi.json). Query latency is advertised as under 100ms.
Capabilities
Use cases
- —Retrieving full documentation for a specific Shopify GraphQL mutation after identifying it via semantic search
- —Fetching complete Stripe API reference pages for an AI coding assistant
- —Loading Cloudflare Workers documentation chunks into an agent's context window
- —Getting detailed Next.js or Tailwind CSS docs to answer developer questions
- —Building agentic workflows that need authoritative, sourced documentation
Fit
Best for
- —AI agents that need structured documentation retrieval across multiple platforms
- —Coding assistants that require full-page context after an initial semantic search
- —Workflows where per-page scraping is too slow or expensive compared to pre-indexed lookups
Not for
- —Searching documentation by meaning (use the /query endpoint instead)
- —Accessing documentation for platforms not yet indexed by AgentUtil
- —Bulk downloading entire documentation sets
Quick start
curl -X POST https://docs.agentutil.net/lookup \
-H "Content-Type: application/json" \
-d '{
"platform": "shopify",
"page": "productVariantCreate"
}'Example
Request
{
"page": "productVariantCreate",
"type": "mutation",
"platform": "shopify"
}Response
{
"page": "productVariantCreate",
"chunks": [
{
"content": "The productVariantCreate mutation creates a new product variant...",
"section": "Overview",
"source_url": "https://shopify.dev/docs/api/admin-graphql/2024-01/mutations/productVariantCreate"
},
{
"content": "Arguments: input (ProductVariantInput!) ...",
"section": "Arguments",
"source_url": "https://shopify.dev/docs/api/admin-graphql/2024-01/mutations/productVariantCreate"
}
],
"platform": "shopify",
"chunks_returned": 2
}Endpoint
Quality
Full OpenAPI schema with request/response definitions and clear pricing from the landing page. However, the x402 challenge was not directly captured (endpoint returned 400 instead of 402), so liveness via the x402 protocol is not confirmed. No standalone docs pages exist beyond the landing page. Response example is inferred from the schema, not observed.
Warnings
- —Probe did not capture a 402 x402 challenge — endpoint returned 400 on POST, possibly requiring a valid JSON body to trigger payment flow
- —No dedicated documentation pages found (/docs, /api, /pricing all return 404)
- —Response example is inferred from OpenAPI schema, not from an actual observed response
Citations
- —The /lookup endpoint retrieves a specific documentation page by ID and returns full chunked contenthttps://docs.agentutil.net
- —Pricing is $0.002 per request via x402 protocol using USDC on Basehttps://docs.agentutil.net
- —Free tier provides 10 queries/day per IP with no signuphttps://docs.agentutil.net
- —~38K chunks across 10 platforms indexedhttps://docs.agentutil.net
- —Indexed platforms include Shopify, Stripe, Cloudflare Workers, Next.js, Anthropic API, OpenAI API, Tailwind CSS, htmx, x402, and Jessehttps://docs.agentutil.net
- —Queries resolve in under 100mshttps://docs.agentutil.net
- —OpenAPI spec defines required fields platform and page with optional type filterhttps://docs.agentutil.net/openapi.json