Skip to main content

​
Overview

Weaver provides multiple health check endpoints to monitor service availability and system status. These endpoints are essential for load balancers, orchestration systems, and monitoring tools.

​
Health Endpoint

GET /health
Basic health check that returns service uptime and status. This endpoint always returns 200 OK when the server is running.

​
Response Fields

​
status
string
Always returns "ok" when the service is running
​
uptime
string
Service uptime duration (e.g., "2h15m30s")

​
Example Request

curl http://localhost:8080/health

​
Example Response

{
  "status": "ok",
  "uptime": "2h15m30s"
}

​
Ready Endpoint

GET /ready
Readiness check that indicates whether the service is ready to accept requests. This endpoint performs health checks and returns 503 Service Unavailable if the service is not ready.

​
Response Status Codes

  • 200 OK - Service is ready to accept requests
  • 503 Service Unavailable - Service is not ready (starting up or failing health checks)

​
Response Fields

​
status
string
Either "ready" or "not ready"
​
uptime
string
Service uptime duration (e.g., "2h15m30s")
​
checks
object
Map of health check results, each containing:
  • name (string) - Name of the health check
  • status (string) - Either "ok" or "fail"
  • message (string) - Optional status message
  • timestamp (string) - ISO 8601 timestamp of the check

​
Example: Service Ready

curl http://localhost:8080/ready

{
  "status": "ready",
  "uptime": "2h15m30s",
  "checks": {
    "database": {
      "name": "database",
      "status": "ok",
      "message": "Connected",
      "timestamp": "2026-03-01T10:30:00Z"
    },
    "agent_loop": {
      "name": "agent_loop",
      "status": "ok",
      "message": "Running",
      "timestamp": "2026-03-01T10:30:00Z"
    }
  }
}

​
Example: Service Not Ready

curl http://localhost:8080/ready

HTTP/1.1 503 Service Unavailable

{
  "status": "not ready",
  "checks": {
    "database": {
      "name": "database",
      "status": "fail",
      "message": "Connection timeout",
      "timestamp": "2026-03-01T10:30:00Z"
    }
  }
}

​
Admin Status Endpoint

GET /admin/status
Returns detailed system status including internal metrics and agent loop information. This endpoint should be protected in production environments.

​
Response Fields

​
uptime
string
Service uptime duration
​
ready
boolean
Whether the service is ready to accept requests
​
...
various
Additional system metrics provided by the agent loop (memory usage, active sessions, queue depth, etc.)

​
Example Request

curl http://localhost:8080/admin/status

​
Example Response

{
  "uptime": "2h15m30s",
  "ready": true,
  "active_sessions": 5,
  "queue_depth": 2,
  "memory_usage_mb": 512,
  "goroutines": 47
}

​
Admin Service Endpoint

GET /admin/service
Returns systemd service status (Linux only). Executes systemctl status weaver and returns the output.

​
Response Fields

​
output
string
Raw output from systemctl status weaver command
​
error
string
Error message if the command failed

​
Example Request

curl http://localhost:8080/admin/service

​
Example Response

{
  "output": "● weaver.service - Weaver AI Agent\n   Loaded: loaded (/etc/systemd/system/weaver.service; enabled)\n   Active: active (running) since Sun 2026-03-01 08:00:00 UTC; 2h ago\n   Main PID: 1234 (weaver)\n   Tasks: 47\n   Memory: 512.0M\n   CGroup: /system.slice/weaver.service\n           └─1234 /usr/local/bin/weaver serve"
}

​
Admin Logs Endpoint

GET /admin/logs
Returns recent service logs (Linux only). Executes journalctl -u weaver -n 100 --no-pager to retrieve the last 100 log entries.

​
Response Fields

​
output
string
Raw output from journalctl command
​
error
string
Error message if the command failed

​
Example Request

curl http://localhost:8080/admin/logs

​
Example Response

{
  "output": "Mar 01 10:30:00 server weaver[1234]: INFO Starting agent loop\nMar 01 10:30:01 server weaver[1234]: INFO Health server listening on :8080\nMar 01 10:30:15 server weaver[1234]: INFO Received chat request session=user-123\n..."
}

​
Integration with Orchestration

​
Kubernetes Liveness Probe

livenessProbe:
  httpGet:
    path: /health
    port: 8080
  initialDelaySeconds: 30
  periodSeconds: 10

​
Kubernetes Readiness Probe

readinessProbe:
  httpGet:
    path: /ready
    port: 8080
  initialDelaySeconds: 10
  periodSeconds: 5
  failureThreshold: 3

​
Docker Compose Health Check

services:
  weaver:
    image: weaver:latest
    healthcheck:
      test: ["CMD", "curl", "-f", "http://localhost:8080/health"]
      interval: 30s
      timeout: 10s
      retries: 3
      start_period: 40s

​
Implementation Details

The health endpoints are implemented in /home/daytona/workspace/source/pkg/health/server.go and provide:
  • /health at line 233 - Basic liveness check
  • /ready at line 246 - Readiness check with health validations
  • /admin/status at line 136 - Detailed system metrics
  • /admin/service at line 147 - Systemd service status
  • /admin/logs at line 159 - Recent service logs

​
Registering Custom Health Checks

Health checks can be registered programmatically using the RegisterCheck method: 
server.RegisterCheck("database", func() (bool, string) {
    if db.Ping() == nil {
        return true, "Connected"
    }
    return false, "Connection failed"
})
Registered checks are automatically included in the /ready endpoint response.

Build docs developers (and LLMs) love

Get started for freeTalk to us