Infrastructure · For agents

The trust & readiness registry for the agent web

Before an agent reads, cites, or transacts with a website, it needs one thing humans take for granted: is this domain trustworthy and ready? aater is a neutral, cross-platform registry that answers that — independent of any single model, CDN, or marketplace.

What an agent can ask

Every domain aater has checked carries machine-readable structural gates: Reachability (can it be fetched without executing JS), Legibility (is the content structured and extractable), and Content Type (informational or commercial — read from declared schema, then URL and structure). Plus, where Pulse is installed, observed agent activity — the part that actually matters.

Reachable?
Fetchable without a headless browser
Legible?
Structured, extractable content
Content type?
Informational or commercial

Query it over MCP or HTTP

aater speaks MCP, so any compliant agent can pull a domain’s structural gates as a tool call — no scraping, no guessing:

# Model Context Protocol endpoint
https://aater.ai/api/mcp

# Or read a domain's classification over HTTP
GET https://aater.ai/api/v1/classify/{domain}

See the developer docs for keys, schema, and rate limits.

Why neutral matters

A registry owned by one model vendor or one CDN is not a registry an agent can trust to be impartial. aater is Switzerland: cross-platform, vendor-neutral, and honest about what it can and can’t prove — observed activity is labeled observed, structural gates are labeled checkable facts, never false certainty about citation.

Verifiable by design

Agent activity aater observes is sealed into a tamper-evident audit log — so the registry isn’t just a claim, it’s a record you can export and independently verify.

Building an agent that needs to know which sites to trust?

Read the developer docs →