Skip to content

Hybrid Search on AlloyDB

Guides an agent through designing, validating, and deploying hybrid (semantic plus keyword) search on Google Cloud AlloyDB.

Repository
google/skills
Stars
21,782
Licence
Apache-2.0
Files
6 files

What it does

When installed, the agent follows a structured workflow for building a search system that combines meaning-based (vector) search with keyword and attribute filtering, using AlloyDB and related Google Cloud services. It first asks detailed questions about your data, performance, security, cost and reliability needs, then produces an architecture, validation plan and deployment guidance, grounding its advice in the bundled Google product mapping and design references and in Google's documentation.

Great for

  • Engineers building product or catalog search on Google Cloud who want a documented architecture and deployment plan
  • Teams evaluating whether AlloyDB is a fit for combining vector search with SQL filtering and reranking

Skip it if

  • Simple keyword-only search, which the skill itself says it should not be used for
  • Projects that require a standalone non-relational vector database, since the skill states it is not the right fit there
  • Users wanting a quick answer without a multi-step requirements interview, because the workflow halts for user input at several points

Try asking

  • “Help me design a hybrid search system for our apparel catalog using AlloyDB with color and price filters”
  • “Create an architecture and deployment plan for semantic plus keyword product search with a p95 latency target under 100ms”
  • “Validate a proposed AlloyDB vector search design and generate test scripts for recall evaluation”
Step-by-step guide: how to use Hybrid Search on AlloyDBInstall, first run, examples, tips and FAQs

Install in 3 steps

  1. 1

    Download the skill

    The ZIP contains the 6 original files, the Apache-2.0 licence and an attribution note.

    google-cloud-solution-hybrid-search-alloydb.zip
  2. 2

    Add it to your agent

    Claude Code (all projects)

    mkdir -p ~/.claude/skills && unzip google-cloud-solution-hybrid-search-alloydb.zip -d ~/.claude/skills/

    Claude Code / Agent SDK (this project only)

    mkdir -p .claude/skills && unzip google-cloud-solution-hybrid-search-alloydb.zip -d .claude/skills/

    Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.

  3. 3

    Ask for the task

    No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.

Get better results

  • Provide your catalog size, search modalities, latency and recall targets up front so the agent does not need to re-ask them
  • Review any gcloud or API commands and validation scripts before allowing execution, since the skill generates and runs them against your Google Cloud environment
  • Use a sandbox or non-production project for the validation phase

What's inside

  • assets/output-template.md
  • references/attribution.md
  • references/design-recommendations.md
  • references/product-mapping.md
  • references/related-guidance.md
  • SKILL.md

Redistributed unmodified from google/skills under the Apache-2.0 licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.