Skip to main content

​
Overview

Monitor your agents through the dashboard or CLI to track performance, debug issues, and optimize execution.

​
Viewing Logs

​
Via Dashboard

1

Go to Runs

Navigate to Agents → my_agent → Runs
2

Select a run

Click on any run to view details
3

View logs

See execution timeline, logs, and results
What you’ll see:
  • Real-time execution logs
  • Action calls and results
  • Memory operations
  • Integration API calls
  • Errors and warnings
  • Final result
  • Execution time

​
Via CLI

# Stream logs for a specific run
ziet logs run_abc123

# Follow logs in real-time
ziet logs run_abc123 --follow

# View recent logs for an agent
ziet logs my_agent --limit 100

# Filter by level
ziet logs my_agent --level error
ziet logs my_agent --level warning

# Filter by date
ziet logs my_agent --since "1 hour ago"
ziet logs my_agent --since "2024-01-15"

​
Via API

curl https://api.ziet.ai/runs/run_abc123/logs \
  -H "Authorization: Bearer YOUR_API_KEY"
Response: 
{
  "logs": [
    {
      "timestamp": "2024-01-15T10:30:00.123Z",
      "level": "info",
      "message": "Starting agent execution"
    },
    {
      "timestamp": "2024-01-15T10:30:01.456Z",
      "level": "info",
      "message": "Calling action: search_flights"
    },
    {
      "timestamp": "2024-01-15T10:30:05.789Z",
      "level": "info",
      "message": "Memory added: key=search_results"
    }
  ]
}

​
Metrics

​
Agent Performance

View metrics in the dashboard for any agent: Execution Metrics:
  • Average execution time
  • P50, P95, P99 latency
  • Success rate
  • Error rate
Resource Metrics:
  • Action calls per run
  • Memory operations (reads/writes)
  • Integration API calls
  • Memory usage
Volume Metrics:
  • Total runs
  • Runs per day/hour
  • Active users

​
Example Dashboard View

my_agent
├── Success Rate: 98.5%
├── Avg Execution Time: 12.3s
├── Total Runs: 1,234
├── Error Rate: 1.5%
└── Last 24h: 87 runs

​
Run Status

​
Check Run Status

  • CLI
  • API
  • Python SDK
ziet status run_abc123
Output:
Run: run_abc123
Agent: my_agent
Status: completed
Started: 2024-01-15 10:30:00
Duration: 12.3s
Result: {"flights_found": 42}

​
Run Statuses

  • running - Currently executing
  • completed - Finished successfully
  • failed - Encountered an error
  • cancelled - Manually cancelled
  • timeout - Exceeded time limit

​
Recent Runs

​
List Runs

  • CLI
  • Python SDK
  • Dashboard
# List recent runs
ziet runs my_agent --limit 10

# Filter by status
ziet runs my_agent --status failed
ziet runs my_agent --status completed

# Filter by date
ziet runs my_agent --since "1 day ago"

​
Webhooks

Get notified when agents complete or fail.

​
Setup Webhook

1

Add webhook in dashboard

  1. Go to Settings → Webhooks
  2. Click Add Webhook
  3. Enter webhook URL
  4. Select events: run.completed, run.failed, run.started
  5. Click Save
2

Receive webhook

Your endpoint receives POST requests:
{
  "event": "run.completed",
  "run_id": "run_abc123",
  "agent_id": "my_agent",
  "status": "completed",
  "result": {
    "flights_found": 42
  },
  "timestamp": "2024-01-15T10:30:00Z"
}
3

Verify signature

import hmac
import hashlib

def verify_webhook(payload, signature, secret):
    expected = hmac.new(
        secret.encode(),
        payload.encode(),
        hashlib.sha256
    ).hexdigest()
    
    return hmac.compare_digest(signature, expected)

​
Webhook Events

  • run.started - Agent run started
  • run.completed - Agent run completed successfully
  • run.failed - Agent run failed
  • run.cancelled - Agent run was cancelled

​
Example Webhook Handler

from flask import Flask, request
import hmac
import hashlib

app = Flask(__name__)
WEBHOOK_SECRET = "your_webhook_secret"

@app.route('/webhook', methods=['POST'])
def handle_webhook():
    payload = request.data
    signature = request.headers.get('X-Ziet-Signature')
    
    # Verify signature
    if not verify_signature(payload, signature):
        return 'Invalid signature', 401
    
    data = request.json
    event = data['event']
    
    if event == 'run.completed':
        print(f"Run {data['run_id']} completed!")
        # Process result...
    
    elif event == 'run.failed':
        print(f"Run {data['run_id']} failed!")
        # Send alert...
    
    return 'OK', 200

def verify_signature(payload, signature):
    expected = hmac.new(
        WEBHOOK_SECRET.encode(),
        payload,
        hashlib.sha256
    ).hexdigest()
    return hmac.compare_digest(signature, expected)

​
Alerts

​
Email Alerts

Get notified via email when agents fail:
  1. Go to Settings → Alerts
  2. Enable Email Alerts
  3. Select events: run.failed, run.timeout
  4. Add email addresses
  5. Save

​
Slack Alerts

Send notifications to Slack:
  1. Go to Settings → Integrations → Slack
  2. Connect your Slack workspace
  3. Go to Settings → Alerts
  4. Enable Slack Alerts
  5. Select channel and events
  6. Save

​
Best Practices

Check your agent’s error rate regularly:
# View failed runs
ziet runs my_agent --status failed --limit 10

# Check error logs
ziet logs my_agent --level error
Get notified immediately when important agents fail:
  • Settings → Webhooks
  • Subscribe to run.failed events
  • Integrate with your alerting system
Check execution time trends:
  • Dashboard → Agents → my_agent → Metrics
  • Look for sudden increases in execution time
  • Optimize slow actions
import logging

logger = logging.getLogger(__name__)

# Info for normal operation
logger.info("Processing started")

# Warning for concerning but non-fatal
logger.warning("Rate limit approaching")

# Error for failures
logger.error("API call failed")

​
Debugging Failed Runs

​
View Error Details

# Get run details
ziet status run_abc123

# View full logs
ziet logs run_abc123

# View error logs only
ziet logs run_abc123 --level error

​
Common Issues

Cause: Agent exceeded time limitSolution: Increase timeout or optimize actions
@Action(
    id="slow_action",
    name="Slow Action",
    description="...",
    timeout=300  # 5 minutes
)
def slow_action():
    ...
Cause: API call to integration failedSolution: Check logs for API errors, verify API keys
# Check integration status
ziet logs run_abc123 | grep "integration"
Cause: Missing memory keySolution: Ensure actions store data before retrieval
# Store first
memory.add(key="data", value=result)

# Then retrieve
data = memory.get("data")

​
Next Steps

Deploying

Deploy your agents

Invoking

Run your agents

API Reference

Full API documentation

Memory

Debug memory operations