How we build these pages
Where the data comes from, how pages stay fresh, and who reviews them.
Every page under /apis is an entry in AgentCash's directory of payment-protected APIs that AI agents can call directly -no API keys, pay per call in USDC on Base.
Each page combines live structured data from the provider's published OpenAPI spec with a short, human-reviewed overview written by AgentCash.
Where the data comes from
Endpoint names, HTTP methods, parameters, authentication modes, and per-call prices are read directly from each provider's published OpenAPI specification. Nothing on a live page describes a capability that isn't present in that spec.
When a page is generated, we capture a fresh snapshot of the provider's endpoints at that moment. The snapshot timestamp is visible as the "Updated" badge on every page.
How categorization works
For each API, we pick the single broadest developer-intent category it belongs to -terms like enrichment, web scraping, image generation, or social media data. Categories are never a vendor's brand name; they reflect what a developer is searching for, not a provider's marketing.
Every category URL ends in -for-ai-agents because that's who these APIs are for.
How pages are reviewed
Overview copy -the headline, summary, and FAQ on each page -is drafted against a strict schema tied to the live OpenAPI snapshot, and then reviewed by a human before publishing. Nothing goes live automatically.
How we keep it fresh
When a provider changes their OpenAPI spec -adds an endpoint, changes a price, deprecates a route -we re-snapshot the provider and regenerate the page. Regenerated pages go back through human review before republishing.