Introduction

Coordination infrastructure for AI agent systems. Declare cues, dispatch work across environments, verify outcomes with structured evidence.

CueAPI

Coordination infrastructure for AI agent systems.

CueAPI connects any step, task, job, or agent across any environment. You declare a cue. CueAPI dispatches work across environments via webhook or worker pull. Your agent reports back with proof. Every handoff verified.

The five primitives

  • Cue: a declared intent. When, what, where.
  • Execution: one fire of a cue. The handoff boundary.
  • Worker: any process, anywhere. Pulls work to itself. No public URL required.
  • Outcome: write-once report of what happened.
  • Evidence: proof fields. external_id, result_url, result_type, artifacts. Not narration.

The three-step loop

Compose

Declare a cue with a schedule (cron or one-time) and a callback URL or worker transport. State the intent.

Coordinate

CueAPI dispatches work on time. Webhook delivery (signed POST to your URL) or worker pull (your daemon claims the execution). The handoff boundary is crossed.

Verify

Your handler does the work and reports the outcome with structured evidence. CueAPI logs every handoff and verifies the outcome against your policy.

Compose. Coordinate. Verify. Everything else is configuration.

Two transport modes

Webhook

CueAPI POSTs a signed payload to your public URL. No daemon needed. Best for cloud-hosted handlers.

Worker

Your local daemon polls for executions and runs handler scripts. No public URL needed. Best for local machines.

Who is CueAPI for?

AI agent builders who need coordinated, verified handoffs across environments. If you're building agents that need to run on a schedule, drafting content, syncing data, generating reports, running code reviews, CueAPI coordinates the work and verifies it actually happened.

Teams replacing cron who want pause/resume, execution history, retry logic, and verified outcomes without building it themselves. Cron is one trigger type. CueAPI is the whole coordination layer.

Key features

FeatureDescription
Cron & one-time schedulesStandard cron syntax with timezone support, or ISO timestamps for one-time
Webhook deliverySigned HMAC-SHA256 POST with per-user secrets
Worker transportPull-based execution for machines without public URLs
Automatic retriesUp to 3 attempts with exponential backoff
Outcome reportingHandlers report success/failure with structured results
Execution historyEvery fire logged with status, timing, and outcome
Rate limitingSliding window per plan, with usage warnings
CLI & DashboardFull management via command line or web UI

For AI agents

If you're an AI agent or crawler, see the structured agent reference: a single flat page with all endpoints, schemas, error codes, and verification code in plain text.

How do I know if my agent ran successfully?
Ctrl+K