Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.sequentum.com/llms.txt

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

The Sequentum Cloud API is REST over HTTPS. Requests are JSON; responses are JSON. Every resource lives under /api/v1 on the production base URL: 
https://dashboard.sequentum.com/api/v1

​
Conventions

  • Authentication — Authorization: ApiKey {key} header. See Authentication.
  • Identifiers — Agents use integer agentId. Runs use integer runId. Sessions use string sessionId.
  • Times — ISO 8601, UTC. Always serialized as strings.
  • Pagination — list endpoints that paginate accept pageIndex (1-based) and recordsPerPage. When provided, the response is { items: [...], totalRecordCount: N }; when omitted, the response is a plain array (backward-compatible shape on legacy endpoints). Endpoints that don’t paginate return the full list.
  • Rate limits — 100 requests per 60-second window per API key, with X-RateLimit-Limit, X-RateLimit-Window-Seconds, and Retry-After (on 429) response headers. See Errors.
  • Errors — RFC 7807 problem details. See Errors.

​
Resources

Agent Builder

Conversational sessions that produce Agents.

Agents

List, run, version, archive — the core Agent lifecycle.

Schedules

Cron, run-once, and interval triggers attached to Agents.

Analytics

Run summaries, records summaries, and per-run diagnostics.

Billing

Credits balance, spending, history, and per-agent cost breakdowns.

Spaces

Workspaces grouping agents, members, and credits.

​
Choosing a transport

Most endpoints are simple request/response. A few patterns are worth knowing:
  • Synchronous vs. asynchronous run start — POST /api/v1/agent/{agentId}/start accepts an isRunSynchronously flag. When false (default), the call returns immediately with a runId and you poll for completion. When true, the call blocks until the run finishes and returns extracted records inline. Use sync only when the run is fast and you need the data in the same request — async is the right default for production.
  • Agent Builder events — GET /api/v1/agent-builder/{sessionId}/events is designed for long-polling with a lastEventIndex cursor. See Sessions.
  • Polling run status — for async runs, poll GET /api/v1/agent/{agentId}/run/{runId}/status until the status is terminal: Completed (9), Success (10), Failure (6, restartable), Stopped (8), or Skipped (11). See Runs for the full lifecycle.

​
Tools

​
SDKs

Official SDKs are coming for Node and Python. Until then, every endpoint works fine over curl or fetch / requests.