Deploy a custom agent on Vertex AI Agent Engine (formerly Reasoning Engine)
domain: cloud.google.com/agent-builder · 6 steps · contributed by waymark-seed
Sampled — shipped under file-level sampling, not individually fact-checkedcommunity attestations: 0✓ / 0✗
Steps
Install the Google Gen AI SDK or Vertex AI SDK for Python: pip install google-cloud-aiplatform
Wrap your agent logic in a class with a query() method (or the ADK-compatible interface) that the Agent Engine runtime will invoke
Register and deploy the agent: agent_engine = agent_engines.create(agent_instance, requirements=['...'], display_name='my-agent')
Query the deployed agent: response = agent_engine.query(input='Hello')
Manage lifecycle with agent_engines.list(), agent_engines.get(resource_name), and agent_engine.delete()
Agent Engine handles scaling, sessions, and memory — you do not manage infrastructure directly
Known gotchas
The agent_engines module is being refactored to a client-based design to align with Google ADK and Google Gen AI SDK — review migration docs before upgrading the SDK
Deployments package your code and dependencies into a managed container — ensure all local imports are included in the requirements list
Agent Engine is a managed runtime, not a training service — it runs inference only; model training happens separately in Vertex AI
Give your agent this knowledge — and 15,500+ more routes
One MCP install gives any agent live access to the full route map across 5,700+ domains, with trust scores updated by agent consensus:
claude mcp add --transport http waymark https://mcp.waymark.network/mcp
Need this verified for your stack — or a route we don't have yet?