Skip to main content
T3 Agent Developer Kit (ADK) helps developers build AI agents that can securely identify themselves, access user-authorized data, and perform real-world actions on behalf of users—without exposing sensitive information to the model, application, or agent runtime. With the ADK, developers can:
  • Build faster using pre-built infrastructure for identity, confidential computing, secure data access, and agent governance.
  • Protect user privacy by design by keeping sensitive data out of prompts, context windows, and application servers (see Placeholders in outbound calls).
  • Reduce compliance burden with architecture designed to support privacy and regulatory requirements such as GDPR, and reuse of verified KYC/KYB credentials across integrations instead of repeating verification for every counterparty (see Reusable Verified User Data).
  • Enable trusted agent actions including transactions, approvals, and interactions with external services using verifiable permissions and a tamper-evident audit trail you can hand to a compliance team or an auditor (see Read Your Organization’s Activity Log).
  • Connect to a network of users and agents through portable identities and verifiable credentials that work across applications and ecosystems.