How To Guides¶
Step-by-step guides for Trust3 AI Governance workflows — from ingesting AI asset inventory to governance processes and external integrations.
Run the AI Assets Collector¶
The AI Assets Collector collects AI-related assets from connected platforms, builds an inventory, and sends relationships into AI Governance so you get a unified view of agents, identities, and posture.
| Step | What to do |
|---|---|
| 1 | Download docker-compose.yml into a directory on your machine (for example ai-assets-collector/). |
| 2 | Create .env in that same directory with DATABRICKS_*, AZURE_*, and AI_GOVERNANCE_API_KEY as needed for your deployment. |
| 3 | Run docker compose up -d, then docker compose logs -f trust3-ai-assets-collector to verify the collector. While the stack stays up, collection repeats every 15 minutes (current release). Run docker compose down when you no longer want scheduled collection on that host. |
Full guide: Run Collector (download link, variable reference table, and checklist).
Platform prerequisites:
- Databricks — workspace URL, personal access token, environment naming
- Azure — Microsoft Entra app registration, secrets, Power Platform app user, least-privilege roles
- Azure Foundry — Microsoft Foundry assign roles to the service principal
Collector hub: AI Assets Collector — one-page index of collector guides.
Custom ACR with Runtime (AKS)¶
Use this path when your security policy requires runtime images from your own Azure Container Registry instead of the Trust3 image hub. The guide covers Advanced Config, image mirroring, and the privacera-hub-secret pull secret on AKS.
For the default Runtime setup with the Trust3 image hub (Basic Config), see Run Collector (Runtime on K8s tab).
Full guide: Using Custom ACR with Runtime
Pre-development approval workflow¶
Before production deployment, agents can go through a structured Legal and Compliance approval chain with an auditable record. See Pre-Development Approval Workflow for roles, registration steps, workflow counters, and audit evidence.
MCP configuration¶
Connect external AI clients (Cursor, Claude Desktop, custom agents) to Trust3 via the Model Context Protocol (MCP). See MCP Configuration for prerequisites, client setup, and verification steps.
Related documentation¶
- Quick Start — sign up and set up the collector end to end