Diagnoses wrong gradients in differentiable NVIDIA Warp programs by measuring first — comparing autodiff against finite differences on a shrunk reproduction before proposing any fix.
AlloyDB BasicsSkill
Summary
Google's official AlloyDB for PostgreSQL skill — cluster and instance lifecycle from gcloud, plus hard rules on private connectivity, IAM database auth and backup behaviour.
Features
- gcloud lifecycle for AlloyDB clusters, primary instances and read pools
- Enforced guardrails: PSC over public IP, no 0.0.0.0/0, IAM database auth by default
- Connectivity reference covering PSA, PSC, Auth Proxy and language connectors
- Backups, point-in-time recovery and the independent lifecycle of discrete backups
- Terraform and Kubernetes Config Connector deployment paths
Install This Skill
Add this skill to your favorite AI agent in a few steps.
Skill Content
Usage Instructions
Learn how to use this skill with different AI agents.
Example Usage
"Create an AlloyDB cluster and primary instance in us-central1 on our existing VPC, wire it up for private access from Cloud Run, and use IAM database authentication rather than a password."
Description
AlloyDB for PostgreSQL is Google Cloud's PostgreSQL-compatible database with disaggregated compute and storage, read pools that scale horizontally, and AlloyDB AI for vector and hybrid search inside SQL. Standing one up correctly means getting the cluster, the primary instance, the network path and the auth model right in one go — and an agent working from general PostgreSQL knowledge will confidently produce a design that either fails to connect or is quietly wide open.
This skill from Google's own skills repository grounds the agent in the AlloyDB working set: enabling the API, creating a cluster and a primary instance, and then a reference directory covering architecture and regional availability, the four connectivity options (Private IP via PSA or PSC, Public IP, Auth Proxy), backups and point-in-time recovery, vertical and read-pool scaling, Terraform and Config Connector deployment, and IAM roles down to database-user level.
What sets it apart from a documentation dump is the block of standing directives it hands the agent. It must recommend Private Service Connect over Public IP, reject any design with 0.0.0.0/0 in authorised networks, default to IAM database authentication over static passwords, insist on roles/alloydb.client rather than roles/alloydb.admin for application connections, require the Auth Proxy or a language connector instead of a raw TCP connection, and state explicitly that IAM database users cannot be created with SQL alone. Those are the failure modes that show up in real AlloyDB deployments, encoded as rules rather than prose.
Use it when an agent is provisioning, connecting to, or reviewing AlloyDB. It deliberately excludes Cloud SQL and other Google Cloud databases — reach for the Cloud SQL skill for those.
Related Skills
Google's official skill for the gws CLI — drive Gmail, Drive, Calendar, Sheets, Docs, Chat and Admin APIs from an agent, with Model Armor screening.
Netlify's official skill for zero-config managed Postgres — querying from Functions, Drizzle setup, migrations and per-preview database branches.
Official WordPress skill for Gutenberg block work: block.json, attributes and serialization, dynamic rendering, and the deprecation path that keeps existing content valid.
