Schedules

Agents degrade silently. A model update changes behavior. A knowledge base entry goes stale. A tool endpoint returns different data. Nothing in your logs flags it — until a customer does.

Schedules run your scenarios automatically on a recurring basis. Set up a nightly test that runs 20 scenarios against your production agent. If the score drops, you know before your team starts their day.

How Schedules Work

A schedule connects a scenario to a timer. When the timer fires, Chanl executes the scenario — same personas, same scorecard, same grading. Results feed into analytics so you can track score trends over time.

Creating a Schedule

curl -X POST "https://api.chanl.ai/api/v1/schedules" \
  -H "Authorization: Bearer <access_token>" \
  -H "Content-Type: application/json" \
  -d '{
    "scenarioId": "scenario_abc123",
    "frequency": "daily",
    "time": "09:00",
    "timezone": "America/New_York",
    "enabled": true
  }'

const schedule = await fetch('https://api.chanl.ai/api/v1/schedules', {
  method: 'POST',
  headers: {
    'Authorization': `Bearer ${accessToken}`,
    'Content-Type': 'application/json',
  },
  body: JSON.stringify({
    scenarioId: 'scenario_abc123',
    frequency: 'daily',
    time: '09:00',
    timezone: 'America/New_York',
    enabled: true,
  }),
});

Frequency Options

CI/CD Integration

Run scenarios after every deployment. If scores drop below your threshold, fail the pipeline:

# In your deployment script
chanl scenarios execute <scenario-id> --wait

# Check the result
SCORE=$(chanl simulations list --scenario <scenario-id> --latest --json | jq '.averageScore')
if [ "$SCORE" -lt 80 ]; then
  echo "Agent quality dropped to $SCORE — blocking deployment"
  exit 1
fi

Managing Schedules

# List active schedules
chanl schedules list

# Pause a schedule
chanl schedules update <id> --enabled false

# Delete a schedule
chanl schedules delete <id>

What's Next