The lightweight, developer-first safety layer for AI actions
Queue, approve, and execute API calls safely from your agents. Add human approval, retries, and audit trails in minutes.
If your AI can issue refunds, update customer data, or trigger internal APIs — you need a safety layer.
DispatchQ sits between your agent and the real world.
await fetch("https://api.dispatchq.dev/v1/jobs", {
method: "POST",
headers: {
Authorization: `Bearer ${API_KEY}`,
"Content-Type": "application/json",
},
body: JSON.stringify({
url: "https://api.stripe.com/v1/refunds",
method: "POST",
payload: { charge: "ch_3Qx9R2eZ" },
requires_approval: true,
}),
}) Your AI doesn't call Stripe directly. It goes through DispatchQ.
Four steps. Five minutes. No infrastructure to manage.
Your agent wants to do something real: refund a payment, update a CRM record, trigger a deploy.
Instead of calling the API directly, the action goes through DispatchQ. Require approval, dedupe, add retries, or delay execution.
Approve via MCP inside ChatGPT or Claude, via API, or auto-approve safe actions. You decide what needs human oversight.
Guaranteed delivery with signed requests. Full request and response history. Your agent or system can fetch the result anytime.
Because this happens.
Before
// ❌ BEFORE — your AI calls Stripe directly
await stripe.refunds.create({
charge: "ch_3Qx9R2eZ",
})
// No approval. No retries. No audit trail.
// What if the LLM hallucinates the charge ID?
// What if it retries and double-refunds? After
// ✅ AFTER — your AI goes through DispatchQ
await dispatchq.fetch("https://api.stripe.com/v1/refunds", {
method: "POST",
body: { charge: "ch_3Qx9R2eZ" },
requiresApproval: true,
})
// Human approves. DispatchQ executes.
// Signed delivery. Full audit trail. Safe retries. From AI support agents to internal copilots. DispatchQ handles actions so your AI doesn't go rogue.
Refunds, credits, account changes.
→ Require approval before execution
Modify CRM, run admin actions.
→ Audit everything your AI does
Sync systems, call flaky APIs.
→ Retries + durability built in
A complete safety layer in 10 endpoints.
Pause jobs for sign-off. Approve via MCP, API, or auto-approve safe actions.
Exponential backoff from 20s to 1h. Configurable max attempts per job.
Dedupe keys and idempotency headers. No double refunds from agent retries.
Every response stored. Query status, body, and duration. Results outlive sessions.
Every delivery signed with SHA-256. Verify requests are from DispatchQ.
Cron expressions with timezones. Delay jobs by 30s, 5m, 1h, or 1d.
Optionally get notified when a job completes. Simple JSON, not base64.
Works with ChatGPT, Claude, and Cursor. Full CLI for developers.
Your agent can queue actions, request approval, and check results — all through MCP tools. No custom code needed.
Per job, not per message. No surprise bills.
For testing and side projects
Start freeFor real workloads
Get startedFor production systems
Get startedAll plans include approvals, retries, deduplication, HMAC signatures, stored results, MCP server, and CLI.
No credit card required. Get your API key in 30 seconds.