vertex-doctor
Probes, diagnoses and unblocks model access on Google Cloud Agent Platform (formerly Vertex AI) with the vertex-doctor CLI: which publisher models a project and credential can call, in which location, through which surface, and the one command that fixes each failing cell; plans quota requests (QPM plus input/output tokens per minute, the base_model bucket, the endpoint that has rows). Use when an aiplatform.googleapis.com call returns 401, 403, 404, 429 or 400, when onboarding a project, an SDK, an agent harness or a gateway to Claude, Gemini, Grok, GPT-OSS, Llama, Mistral or embedding models, when asked which models a project can call and where, when a gateway health check reports Vertex models unhealthy, or when someone needs to know what quota to ask for. Don't use for deploying self-hosted Model Garden models, for Agent Gateway / Agent Registry troubleshooting, or for non-Google model providers.
