TERMS OF SERVICE

1. Acceptance of Terms

By accessing or using the AgentOS platform, Python SDK, TypeScript SDK, or REST APIs, you agree to be bound by these Terms of Service. If you are entering into this agreement on behalf of a company or legal entity, you represent that you have the authority to bind such entity.

2. Autonomous Agent Execution & Governance

AgentOS provides an infrastructure runtime to deploy, monitor, and govern autonomous AI agents:

• Sovereign Wallet Spending: You are solely responsible for setting appropriate Layer 4 budget limits, approval thresholds, and funding amounts for your agents.
• Tool Execution Sandboxing: You agree not to register or invoke tools designed for malicious cyber operations, unconsented network port scanning, or illegal data harvesting.
• Human-in-the-Loop (HITL): You acknowledge that transactions exceeding designated approval thresholds require manual human review before execution.

3. Acceptable Use & Prompt Injection Policy

You agree not to bypass Layer 8 Prompt Injection Defenses, disable security telemetry, or use AgentOS agents to generate harmful, fraudulent, or infringing material. Any detected attempts to compromise platform infrastructure or multi-tenant boundaries will result in immediate termination of API tokens.

4. Service Availability & SLA

We target high reliability across the AgentOS API gateway and PostgreSQL poolers. Scheduled maintenance windows are communicated in advance via the public Status page.