FleetPulse quickstart

Connect your first agent

Send a sample health record, find it in your fleet, then review a per-agent remediation policy. These steps describe the signed-in prototype flow available today.

Start 14-day trial Open Fleet dashboard

Current connection boundary

POST /api/agents requires a signed-in FleetPulse session. There is no agent credential or unattended-agent authentication flow yet. Fleet metrics and policies are not scoped per account, and the fleet read endpoint is unauthenticated. Use only non-sensitive sample data; do not send production telemetry or enable remediation for a production fleet until agent credentials and ownership isolation are available.

Setup

From first metric to policy

1. Create a trial account
Start the 14-day trial, then keep the FleetPulse session signed in for the prototype ingestion request.
2. Send a sample metric
From an authenticated FleetPulse session, send this JSON body to POST /api/agents. The values and agent ID are illustrative.
POST /api/agents · example body
```
{
  "agentId": "checkout-worker",
  "latencyMs": 184,
  "tokenUsage": 1260,
  "errorCount": 0,
  "status": "active"
}
```
3. Find the agent in Fleet
Open the Fleet page to review the latest status and the agent's Policy card. Signed-out visitors are sent to sign in.
Go to Fleet
4. Review its remediation policy
The Fleet page loads policy settings for each agent. This example matches the current defaults; optional threshold and budgetPerRun are omitted.
The request body is sent to PUT /api/agents/checkout-worker/policy.
PUT /api/agents/:agentId/policy · example body
```
{
  "restartFailureThreshold": 3,
  "rollbackDowntimeMinutes": 5,
  "enabled": false,
  "action": "restart",
  "cooldownMinutes": 60
}
```
5. Enable only after review
Remediation stays off in the example. Review the action and limits in the Policy card, then turn on its enable switch only when you intend to allow remediation.

Ready to explore with sample data?

Start your trial