Logs

List what AI agents did through the MCP server

One entry per JSON-RPC message that reached the MCP server with a credential of the account — the tool the agent called, through which app (ChatGPT, Claude…), from where, how it ended and which API calls it made (their mcpCallId in GET /v1/logs is this callId). Kept for 180 days, like the API log. This log never keeps tool results, only their size (the API call a tool makes is in GET /v1/logs, redacted like any call); arguments are stored redacted (GET /v1/logs/mcp/{id}). Requests whose credential could not be identified are not in the log (a known key that is refused — switched off, for example — is). Scope read:logs.

get/logs/mcp

Request

  • Base URL: https://api.oveyon.com/v1
  • URL: https://api.oveyon.com/v1/logs/mcp
  • Auth: HTTP bearer

Query parameters

limitinteger
cursorstring

The next_cursor of the previous page. Cursors from different routes are not interchangeable (400 bad_cursor).

outcome'ok' | 'tool_error' | 'rejected' | 'error' | 'aborted'

How the message ended. Anything else is 400 bad_outcome.

toolstring

The tool name (get_inbound_content); implies tools/call. Malformed is 400 bad_tool.

methodstring

The JSON-RPC method (tools/call, tools/list, initialize…). Malformed is 400 bad_method.

credentialinteger

The id of one credential of the account (an app connection or an API key).

fromstring

Start of the range, ISO 8601 (YYYY-MM-DD or timestamp).

tostring

End of the range, ISO 8601.

Response

Page

next_cursorstring nullable required

null = last page.

Changes

No recorded changes to this operation across all 1 revision of this API.