Vertex AI Agent Engine and Dome

1. 01

   Register an agent and issue its key

   Then store the key in Secret Manager.

   $ dome agents register --name support-triager --gateway prod-gateway
   $ dome agents create-key support-triager --name agent-runtime
   $ dome rules apply support-triager.cedar --agent support-triager --name linear-support-only

2. 02

   Pass the key at deploy

   Agent Runtime's env_vars can reference a Secret Manager secret. The agent reads it from the environment.

   env_vars = {
       "GATEWAY_URL": "https://<gateway-host>/gateways/<gateway-id>",
       "DOME_AGENT_KEY": {"secret": "dome-support-triager", "version": "1"},
   }

3. 03

   Point tools and model at the Gateway

   In ADK, an MCP toolset for tools and LiteLLM for the model. The model name is a pool.

   import os
   from google.adk.models.lite_llm import LiteLlm
   from google.adk.tools.mcp_tool import McpToolset, StreamableHTTPConnectionParams
    
   GATEWAY_URL, DOME_AGENT_KEY = os.environ["GATEWAY_URL"], os.environ["DOME_AGENT_KEY"]
    
   tools = McpToolset(connection_params=StreamableHTTPConnectionParams(
       url=f"{GATEWAY_URL}/mcp",
       headers={"Authorization": f"Bearer {DOME_AGENT_KEY}"},
   ))
   model = LiteLlm(model="openai/claude-opus-5-5", api_base=f"{GATEWAY_URL}/v1", api_key=DOME_AGENT_KEY)

Commands and rules tested against a Dome workspace on October 1, 2026. For anything about Vertex AI Agent Engine itself, see Google Cloud's documentation.

permit (principal, action == Dome::Action::"mcp:discover", resource);
 
permit (principal, action == Dome::Action::"mcp:call", resource is Dome::MCPTool)
when { resource.connection_name == "linear" };
 
forbid (principal, action == Dome::Action::"mcp:call", resource is Dome::MCPTool)
when {
  resource.connection_name == "linear" &&
  resource.tool_name == "save_issue" &&
  !(resource has arguments && resource.arguments has team &&
    resource.arguments.team == "Support")
};