Call Volume
Current periodDecision Split
Current periodAgent Activity
This billing periodGovernance & Usage
Audit Log
| Time | Agent | Tool | Arguments | Status |
|---|---|---|---|---|
| Loading... | ||||
Kill Switch
Instantly block a tool by name. Any call to a blocked tool is rejected before rule evaluation — no delay, no AI processing.
Blocked tools
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 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.
TrustLoop Verified Badge
Add this badge to your product or website to show your users that your AI actions are monitored, governed, and independently verifiable.
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.
Policy Desk
Upload your existing AI policy document. TrustLoop will read it and propose governance rules — you review and approve before anything goes live.
Compliance Reports
Generate a compliance report for any date range. Download as PDF, Word (.docx), CSV, or an evidence pack (ZIP with PDF + CSV + blockchain verification).
Tool Registry
Register the tools your agents use. TrustLoop's MCP server will expose them dynamically. Store credentials here so your agents never hold dangerous API keys directly.
Blockchain Anchors
TrustLoop writes a keccak256 hash of all audit log entries to the Polygon blockchain every hour. Click any transaction to verify it independently on Polygonscan — no trust in TrustLoop required.
- Export your audit log from the Overview tab (↓ Export CSV)
- Identify the records within the anchor's one-hour window (ending at the anchor timestamp)
- Recompute the hash in Node.js or Python:
ethers.keccak256(ethers.toUtf8Bytes(JSON.stringify(records))) - Compare your result to the Batch hash in the table above
- Click the transaction link — on Polygonscan, open More details → Input data to see the same hash written on-chain
id, tenant_id, tool_name, status, created_at — ordered by created_at ascending. JSON.stringify is compact (no spaces or newlines).
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.