General Purpose Template¶
The default template for any agent type. Provides sensible baselines for rate limiting, anomaly detection, and destructive action gating.
Default Configuration¶
| Parameter | Default | Description |
|---|---|---|
daily_action_limit | 200 | Max actions per 24 hours |
anomaly_threshold | 0.7 | Anomaly score threshold for blocking |
cooldown_violations | 3 | Number of violations in 24h before cooldown |
approve_destructive | true | Destructive actions require human approval |
Violation Rules¶
| Rule | Trigger | Severity |
|---|---|---|
anomaly_score_threshold_exceeded | Agent's anomaly score > anomaly_threshold | high |
daily_action_limit_exceeded | 24h action count > daily_action_limit | medium |
violation_cooldown_active | Agent has >= cooldown_violations violations in 24h | high |
Approval Rules¶
| Rule | Trigger |
|---|---|
requires_human_approval | Destructive methods (delete, destroy, purge, drop) when approve_destructive is true |
Configuration Example¶
Higher limits for a trusted production agent:
curl -X PATCH https://vargate.ai/api/dashboard/settings \
-H "Content-Type: application/json" \
-H "X-API-Key: YOUR_KEY" \
-d '{
"policy_template": "general",
"policy_config": {
"daily_action_limit": 1000,
"anomaly_threshold": 0.85,
"cooldown_violations": 5,
"approve_destructive": true
}
}'
Good starting point
The general template is applied by default to new tenants. It provides a reasonable baseline that works for most agent types. Switch to a specialized template when you need domain-specific rules.