Developer documentation

Public API and agent integration

Catalog and pricing data are available read-only. Checkout, orders, and payments are not enabled.

Your name stays on the cover.

Your rights stay in your hands.

Your royalties stay in your pocket.

Last updated:

Live read endpoints

  • GET /api/products
  • GET /api/products/{id}
  • GET /api/services
  • GET /api/services/{id}
  • GET /api/pricing
  • GET /api/availability

Machine-readable files

  • /openapi.json — API contract
  • /feeds/products.json — book catalog
  • /feeds/services.json — publishing plans
  • /.well-known/agent.json — agent capabilities and restrictions
  • /llms.txt and /llms-full.txt — AI discovery guidance

Disabled write endpoints

  • POST /api/purchase-intents returns a non-binding human confirmation link
  • POST /api/checkout/session returns HUMAN_APPROVAL_REQUIRED
  • GET /api/orders/{id}/status returns NOT_IMPLEMENTED
  • No public endpoint creates an order, reserves work, accepts terms, or processes payment

Sandbox

  • GET /api/sandbox/products returns marked test data
  • POST /api/sandbox/checkout/session returns a simulated, non-payable response
  • Sandbox responses can never create real orders or payments

Usage and safety

  • Use canonical IDs and cite the source URL
  • Respect cache headers and reasonable request rates
  • Treat missing values as unknown
  • Review AI Agent Access Terms before automated use

Future production controls

  • Authentication, scoped authorization, rate limiting, durable audit logs, payment provider integration, approval UI, and security review are required before write actions can be enabled