Logfire

Set up Logfire for AI Observability

Triggers

On Alert Received Trigger when a Logfire alert is received via webhook

Actions

Query Logfire Execute a read-only SQL query against Logfire Query API

Instructions

Create a Logfire API key for SuperPlane

1. Open Settings in Logfire.
2. Under ORG: <your-username>, select API Keys.
3. Click New API Key.
4. Enter a key name.
5. Enable these five scopes:
   • Organization scopes: organization:write_channel (required for auto-creating webhook channels)
   • Project scopes: project:read , project:read_token , project:read_alert , and project:write_alert
6. Select All Project or a specific project from the dropdown.
7. Click Create API Key.
8. Copy the API key and paste it.

On Alert Received

The On Alert Received trigger starts a workflow execution when Logfire sends an alert payload to your SuperPlane webhook URL.

Configuration

Select the Logfire Project and Alert you want to trigger the workflow.

Webhook setup

After you save this trigger, SuperPlane provides a webhook URL. Add that URL as a Logfire notification webhook target so alert events are sent to this workflow.

Example Data

{
  "data": {
    "alertId": "alt_123",
    "alertName": "Latency spike",
    "eventType": "firing",
    "message": "p95 latency exceeded threshold",
    "severity": "warning",
    "url": "https://logfire-us.pydantic.dev/my-org/my-project/alerts/alt_123"
  },
  "timestamp": "2026-03-23T12:00:00.000000000Z",
  "type": "logfire.alert.received"
}

Query Logfire

The Query Logfire component executes a read-only SQL query against Logfire and returns query results for use in downstream steps.

Use Cases

• Investigate traces and spans: Query recent records for errors, latency spikes, or specific services
• Build reporting workflows: Export Logfire data into Slack, email, dashboards, or data stores
• Conditional automation: Query for specific conditions, then branch workflow logic based on returned rows
• Scheduled analytics: Run recurring SQL queries to monitor usage and operational metrics

Configuration

• Project: Required Logfire project to query (scopes the generated read token)
• SQL: Required SQL query (supports expressions). Example: SELECT start_timestamp, message FROM records LIMIT 10
• Time Window: Optional preset time window (e.g., Last 5 minutes, Last 1 hour). Select “Custom” to specify exact timestamps
• Limit: Optional maximum rows to return. If omitted, Logfire defaults to 500; maximum is 10000
• Row Oriented: Optional JSON format toggle. false returns column-oriented JSON; true returns row-oriented JSON

Output

Emits one logfire.query event containing the Logfire query response (for example columns and/or rows, depending on format options).

Use this output to transform, filter, or route query results to other components.

Example Output

{
  "data": {
    "columns": [
      {
        "name": "start_timestamp",
        "type": "timestamp"
      },
      {
        "name": "message",
        "type": "text"
      }
    ],
    "rows": [
      [
        "2026-01-01T00:00:00Z",
        "Example Logfire record"
      ]
    ]
  },
  "timestamp": "2026-03-23T12:00:00.000000000Z",
  "type": "logfire.query"
}