Governance Dashboard
Audit Log
| Time | Agent | Tool | Arguments | Status |
|---|---|---|---|---|
| Loading... | ||||
Kill-Switch Rules
AI-Powered Approval Rules
How AI Rules Work
Write rules in plain English. TrustLoop uses AI to evaluate every tool call against your rules — no need for exact tool names.
✅ "Block wire transfers over $5,000"
✅ "Flag any export containing PII data"
✅ "Route all database write operations for review"
Block immediately — rejects the action with no approval option
Pending Approvals
Actions waiting for your decisionDecision History
Notification Settings
Configure where TrustLoop sends approval requests when an AI agent needs human sign-off.
Global Mandate Coverage
Auto-updated as laws changeRisk Distribution
MEDIUM — send, write, export, payment, or bulk data calls
LOW — read-only, query, list, or informational calls
Data Retention
Automatically purge audit logs older than a set number of days. Required under GDPR Art. 5(1)(e), CCPA, and most national data protection laws to avoid holding personal data longer than necessary.
Data Subject Access Request
Download a full structured export of all data TrustLoop holds for your account. Use this to respond to DSAR requests under GDPR Art. 15, CCPA §1798.110, India's DPDPA §11, or any equivalent legislation.
✅ Format: JSON (machine-readable + human-readable)
✅ PII already masked in stored arguments
Connect your agent
Give your agent a name, pick your integration method, then paste the snippet into your agent code.
Use a unique name per agent — this is how it appears in your audit log dropdown.
- Your agent calls
POST /api/interceptbefore executing each tool. - TrustLoop checks your approval rules and returns
allowed: true/false. - Your agent only runs the tool if TrustLoop says allowed — blocked calls never execute.
- Every decision is logged and visible in your audit log.