Content Discovery¶
Content discovery enables agents to find UCAP-enabled publishers and content sources. UCAP provides two discovery mechanisms: a well-known endpoint for individual publishers and a global registry for cross-publisher discovery.
Well-Known Endpoint¶
Publishers MAY expose a discovery document at a well-known URL on their domain. This allows agents to determine whether a publisher supports UCAP and find the UCAP server endpoint.
Discovery Document Fields¶
| Field | Type | Required | Description |
|---|---|---|---|
ucap_version |
string | Yes | UCAP specification version |
publisher_id |
string | Yes | Publisher's unique identifier |
publisher_name |
string | Yes | Publisher's display name |
ucap_server |
string | Yes | Base URL of the UCAP server |
catalog_url |
string | No | Direct URL to publisher's catalog |
subscribe_url |
string | No | URL for subscription page |
Global Registry¶
UCAP servers MAY maintain a global registry of publishers for agent discovery. The registry allows agents to search for publishers without knowing their individual domains.
Note
The global registry specification is under development. Current implementations should rely on the well-known endpoint for publisher discovery and the Catalog Search capability for content discovery.
Publisher Registration¶
Publishers register with a UCAP server to make their content available to agents.
Registration Requirements¶
| Requirement | Description |
|---|---|
| Content Source | Connection to content platform (API, export, etc.) |
| Tier Configuration | Mapping of content tiers to prices |
| Payment Setup | Payment processor connection (Stripe, etc.) |
Supported Content Sources¶
UCAP is source-agnostic. Common integrations include:
| Source | Integration Method |
|---|---|
| Patreon | OAuth 2.0 + API |
| Ghost | Admin API Key |
| WordPress | REST API + Application Password |
| Substack | Manual export (ZIP upload) |
| Custom | Webhook or direct upload |
Guidelines¶
For Publishers¶
- Publishers SHOULD expose the
/.well-known/ucap.jsondiscovery document - Publishers MUST configure at least one content tier
- Publishers SHOULD provide complete metadata (description, category, media) for catalog listing
- Publishers MAY support multiple content sources
For Agents¶
- Agents SHOULD check
/.well-known/ucap.jsonwhen encountering a publisher URL - Agents SHOULD use the Catalog Search capability for broad content discovery
- Agents SHOULD cache discovery documents for reasonable periods