# auth.md

Agent authentication metadata for Structured Intelligence Ltd,
https://www.structuredintelligence.net

## Summary

**No authentication is required.** Every resource published on this origin is
public and can be fetched by any agent without registration, credentials or
tokens. There is no agent registration endpoint, because there is nothing to
register for.

This file exists to state that explicitly. An agent that finds no auth metadata
cannot tell the difference between a site that needs no credentials and one
whose metadata is simply missing.

## Agent audience

- **Resource identifier:** `https://www.structuredintelligence.net`
- **Audience value for any future token:** `https://www.structuredintelligence.net`
- **Registration required:** no
- **Anonymous access:** yes, for all published resources

## OAuth Protected Resource Metadata

Published per RFC 9728 at:

<https://www.structuredintelligence.net/.well-known/oauth-protected-resource>

It is also advertised in a `Link` header on every response from this origin,
with `rel="oauth-protected-resource"`.

```json
{
  "resource": "https://www.structuredintelligence.net",
  "authorization_servers": [],
  "scopes_supported": [],
  "bearer_methods_supported": ["header"]
}
```

`authorization_servers` and `scopes_supported` are intentionally empty. No
protected API is currently exposed, so there is no issuer to trust and no scope
to request.

## Authorization Server Metadata

Published at:

<https://www.structuredintelligence.net/.well-known/oauth-authorization-server>
 No authorization server
is currently operated for this resource, and the document says so rather than
advertising an issuer that would refuse every request.

## Agent authentication block

- **skill:** none required
- **register_uri:** not applicable, registration is not required
- **identity types supported:** anonymous
- **credential types supported:** none
- **claim URL:** not applicable
- **revocation URL:** not applicable

There is no registration method to complete, because unauthenticated access is
the only access mode offered.

## Available resources, no credentials needed

| Resource | Path | Type |
|---|---|---|
| Site overview for language models | `/llms.txt` | text/plain |
| Site as markdown | `/index.md` | text/markdown |
| SIET product and evidence record | `/siet.md` | text/markdown |
| Other applications | `/platform.md` | text/markdown |
| Planetary risk record | `/global-risk.md` | text/markdown |
| OpenAPI description | `/openapi.json` | application/openapi+json |
| API catalog | `/.well-known/api-catalog` | application/linkset+json |
| Agent descriptor | `/agent.json` | application/json |
| MCP descriptor | `/.well-known/mcp.json` | application/json |
| Agent skills index | `/.well-known/agent-skills/index.json` | application/json |
| Capability manifest | `/.well-known/ai-catalog.json` | application/json |

## Rate limits

None enforced. Please be reasonable: the site is small and the content is
static, so one fetch per document is sufficient.

## Content usage

See `/robots.txt` for Content Signals. In summary: `search=yes`, `ai-input=yes`,
`ai-train=no`. Retrieval, citation and grounding are welcome. The content is not
offered as model training data.

## If this changes

If a protected API is introduced, this file and the OAuth metadata will be
updated together to describe registration, supported credential types and token
acquisition. Until then, treat the absence of credentials as deliberate.

## Contact

For institutional API access, data access or integration enquiries:
hello@structuredintelligence.net, or https://www.structuredintelligence.net/contact