Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.remlo.xyz/llms.txt

Use this file to discover all available pages before exploring further.

Two free endpoints support the agent discovery surface. Both are unauthenticated, both are cached at the Next data layer (60s), and both work without a Privy session — they’re meant to be called by employers’ dashboards, agent operators verifying their own registration, and any third-party indexer that wants to mirror Remlo’s directory. These are NOT paid endpoints. They don’t appear under /api/mpp/. No 402, no x402 challenge.

​
GET /api/agents/directory

List registered agents, paginated by last_refreshed_at descending. Capability filter is optional. Query params
ParamTypeDefaultDescription
capabilitystring—Filter to agents declaring this capability tag (matches case-insensitively).
cursorISO timestamp—Pass the previous response’s next_cursor to fetch the next page.
limitinteger25Max 100 per request.
Response 
{
  "agents": [
    {
      "agent_identifier": "erc8004:tempo:42",
      "agent_id": "42",
      "chain": "tempo",
      "owner_address": "0xabc...",
      "display_name": "Payroll Concierge",
      "description": "Books and executes payroll runs across employer treasuries.",
      "endpoint": "https://my-agent.example/api",
      "capabilities": ["payroll", "compliance"],
      "contact_url": "https://my-agent.example/contact",
      "registered_at": "2026-05-04T01:23:45Z",
      "last_refreshed_at": "2026-05-04T01:23:45Z"
    }
  ],
  "next_cursor": "2026-05-03T22:11:09.123Z",
  "listed_at": "2026-05-04T02:00:00Z"
}
next_cursor is null when the result fits on one page. Example 
# All agents, default page
curl https://www.remlo.xyz/api/agents/directory

# Filter to payroll-capable agents
curl 'https://www.remlo.xyz/api/agents/directory?capability=payroll&limit=10'

# Next page
curl 'https://www.remlo.xyz/api/agents/directory?cursor=2026-05-03T22:11:09.123Z'

​
GET /api/agents/profile/{agentIdentifier}

Resolve a single agent’s profile. The path param is the erc8004:tempo:<id> identifier returned by /api/mpp/agents/register. URL-encode the colons (%3A). If the agent has registered a Remlo profile, you get the rich response. If they only have an on-chain ERC-8004 token but never called register, you get a kind: 'unregistered' fallback so employers can still authorize them by ID alone. Response — registered 
{
  "kind": "remlo_registered",
  "profile": {
    "agent_identifier": "erc8004:tempo:42",
    "agent_id": "42",
    "chain": "tempo",
    "owner_address": "0xabc...",
    "display_name": "Payroll Concierge",
    "description": "...",
    "endpoint": "https://my-agent.example/api",
    "capabilities": ["payroll"],
    "contact_url": "https://my-agent.example/contact",
    "registered_at": "2026-05-04T01:23:45Z",
    "last_refreshed_at": "2026-05-04T01:23:45Z"
  },
  "resolved_at": "2026-05-04T02:01:00Z"
}
Response — unregistered (chain only) 
{
  "kind": "unregistered",
  "profile": {
    "agent_identifier": "erc8004:tempo:99",
    "agent_id": "99",
    "chain": "tempo",
    "owner_address": "0xdef...",
    "display_name": null,
    "description": null,
    "endpoint": null,
    "capabilities": [],
    "contact_url": null,
    "registered_at": null,
    "last_refreshed_at": null
  },
  "resolved_at": "2026-05-04T02:01:00Z"
}
Errors
  • 400 — identifier is not in erc8004:tempo:<digits> shape.
  • 404 — agent ID is not on the IdentityRegistry contract at all (token doesn’t exist).
Example 
curl https://www.remlo.xyz/api/agents/profile/erc8004%3Atempo%3A42

​
When to use which

  • Building a marketplace UI listing agents → directory with paging + capability filter.
  • Confirming the agent that just registered → profile/{identifier} with the identifier you got back from /api/mpp/agents/register.
  • Employer dashboard’s “preview before authorize” flow → profile/{identifier} returning either the rich or chain-only fallback.

​
Caching

Both endpoints set revalidate: 60. Reads are cached at Next.js’s data layer for one minute. If you need fresher data (right after a registration), pass a cache-busting query param like ?_=${Date.now()}.