AegisGate Agent Skill

What problem does it solve?

This Skill helps you install, start, and safely integrate AegisGate on a new machine so your AI applications can route through a security gateway instead of sending raw prompts/responses directly to upstream LLM providers.

Core Features & Use Cases

• Install & boot AegisGate with Docker or run it locally from source, then validate via health/readiness endpoints.
• Choose the correct routing mode (token routing recommended for multi-tenant use, or direct upstream mode for trusted local/agent setups).
• Register and manage gateway tokens so clients can use stable OpenAI-compatible endpoints while ensuring sensitive data handling and policy enforcement.
• Connect AI clients/agents by configuring baseUrl and upstream API keys (including common workflows for OpenAI-compatible, Anthropic Messages, and v2 proxy access patterns).
• Troubleshoot failures with a clear verification order (health → routing mode → token lookup/config → logs).

Quick Start

Ask the agent to install AegisGate with Docker, start it locally, register a token for your upstream (including any whitelist fields you need), and then provide a ready-to-use baseUrl and client config for your LLM toolchain.