PROTOCOL 01
Catalog delivery protocol
Your products are exposed in machine-readable formats so agents can parse offers and availability without scraping templates.
Includes: Feed JSON + markdown agent view
Protocols
Protocol stack
buyers can trust.
Agentsfeed does more than publish endpoints. It ships a complete protocol layer so AI agents can discover your store, understand your catalog, and follow your access rules with minimal friction.
PROTOCOL 02
Well-known discovery protocol
Agents discover your capabilities from standard well-known routes instead of brittle custom integration docs.
Includes: Agent card + skills + API catalog
PROTOCOL 03
Service contract protocol
The API catalog links directly to service description and status so integrators can verify capabilities quickly.
Includes: Linkset catalog + OpenAPI + status
PROTOCOL 04
Auth discovery protocol
OAuth and OpenID metadata are published so authenticated agent workflows can connect with predictable token flows.
Includes: OpenID + OAuth auth server + protected resource
PROTOCOL 05
Policy signaling protocol
Crawler and model policies are explicit, so you control training/input posture and avoid ambiguous bot behavior.
Includes: robots.txt + llms.txt + Content-Signal
PROTOCOL 06
Agent tooling protocol
MCP metadata and WebMCP context expose callable storefront capabilities for compatible AI runtimes.
Includes: MCP server card + browser WebMCP tools
Business impact
Why this matters
for revenue teams.
IMPACT 01
Less integration risk
Standardized discovery reduces one-off implementation errors and helps external teams integrate faster.
IMPACT 02
Faster time to agent traffic
When protocols are already in place, new agent clients can discover your store with less manual onboarding work.
IMPACT 03
More control with less custom code
Policy and protocol behavior lives in module settings, so you can adapt without recurring theme rewrites.
Technical appendix
Full endpoint map,
when you need details.
Implementation teams can use this route inventory for QA and integration checks after setup.
ENDPOINT 01
Product feed
Primary machine-readable catalog resource for product discovery.
/module/agentsfeed/feed
ENDPOINT 02
Agent markdown
Compact markdown summary for lightweight agent crawlers.
/ai/agent-ready.md
ENDPOINT 03
Agent card
Advertises high-level agent identity and capability references.
/.well-known/agent-card.json
ENDPOINT 04
Agent skills
Lists skill/discovery pointers for capable clients.
/.well-known/agent-skills/index.json
ENDPOINT 05
Skill doc
Skill-level markdown docs with stable URLs and digest-backed index entries.
/.well-known/agent-skills/catalog-read.md
ENDPOINT 06
API catalog
RFC linkset catalog with item, service-desc, service-doc, and status relations.
/.well-known/api-catalog
ENDPOINT 07
Service descriptor
OpenAPI service description linked from the API catalog.
/.well-known/service-desc
ENDPOINT 08
Status endpoint
Machine-readable operational status for discovery and monitoring.
/.well-known/status
ENDPOINT 09
OpenID config
OIDC discovery metadata for agent-compatible authentication flows.
/.well-known/openid-configuration
ENDPOINT 10
OAuth auth server
OAuth authorization server metadata for token and grant discovery.
/.well-known/oauth-authorization-server
ENDPOINT 11
OAuth protected resource
Protected resource metadata including authorization servers and scopes.
/.well-known/oauth-protected-resource
ENDPOINT 12
ACP descriptor
Capability descriptor for ACP-aware clients.
/.well-known/acp.json
ENDPOINT 13
UCP descriptor
Capability descriptor for UCP-aware clients.
/.well-known/ucp.json
ENDPOINT 14
MCP server card
MCP server metadata for model context protocol integrations.
/.well-known/mcp/server-card.json
ENDPOINT 15
Robots policy
Crawl and AI policy output controlled by module settings.
/robots.txt
ENDPOINT 16
llms policy
Model-facing policy and index content endpoint.
/llms.txt