Run prompts on Gemini and open cloud models.
Runs approved prompts against hosted models and provides code and troubleshooting guidance for connecting to Google Cloud Agent Platform.
When to use it
Use for model inference, connection code, authentication, endpoint configuration, or inference errors, not model deployment or evaluation.
Give it an inference or connection request; it clarifies missing parameters, obtains approval before inference, and returns generated output with execution details.
Requires Python to run inference scripts through the official SDKs.
Uses the Google Cloud CLI for authentication, API enablement, and endpoint discovery.
This skill
Google Cloud Agent Platform
Sends prompts for cloud inference (irreversible)
docs.cloud.google.com
Checks current model documentation