Index of every page on this site: llms.txt
All skills
google avatar

/product-search

@ead8d7f
by googlegoogle/adk-recipes11k stars
2,961
Run it yourself

Creates product search agents with semantic search and RAG on Google Cloud (Vector Search on Gemini Enterprise Agent Platform, BigQuery, embeddings). Use when the user wants to "build a product search agent", "create an e-commerce search", "make a shopping assistant", "set up semantic catalog discovery", "ingest products into Vector Search", or "deploy a retail RAG agent". Handles the full pipeline: catalog data ingestion to BigQuery, Gemini Enterprise Agent Platform Vector Search collection setup, ADK agent scaffolding, evaluation, and Cloud Run deployment.

  • 28 files
  • 586.9 KB
  • Updated 2 months ago
  • GitHub
Use this Skill: https://skilld.dev/gh/google/adk-recipes/product-search

Nothing lands on disk. Nothing to clean up.

Fork this Skill

Edit a local copy. It keeps the author and licence.

referencesarchitecture.md

≈563 tokens on demand. Your agent reads this file only when SKILL.md points to it.

What the retrieval layer actually does

Read this when demoing the skill or extending it — covers what's semantic vs what isn't, and where structured filtering would have to live.

Semantic search, not filtered search

search(query, top_k) runs Vector Search semantic similarity on Gemini Enterprise Agent Platform over the embedding fields configured in design-spec.md (default: name, description, category, brand). It does NOT apply structured filters on price, currency, stock, or rating -- those words in a query are just text the embedder sees.

Implications when you demo:

  • "headphones under 100 EUR" returns headphones ranked by semantic match; any price filtering happens in the agent's LLM, not the retriever, so results may include items above the threshold (the LLM may then narrate "no matches" if it filters client-side).
  • The bundled sample-products.csv has USD prices. If the user asks in EUR, the agent silently treats them as equivalent. For a real multi-currency demo, add a price_eur or currency column to the catalog and prompt the agent to use it.

Test queries

Query What you're testing
"laptop for video editing" Pure semantic match (works as documented)
"I need a gift" Vague-query clarification flow (agent prompt-driven)
"Which one has the best battery life?" RAG-style follow-up
"wireless headphones under $100" Semantic + LLM-side price filtering; agent will narrate the price filter even though the retriever doesn't apply it

Where to add structured filtering

If a real catalog needs price < X to actually constrain results (not just narrate around them), there are three places to add it:

  1. Inside the retriever — pre-filter the BigQuery dataset before embedding ingest, or post-filter the Vector Search results in scripts/retrievers.py:search().
  2. As a separate tool on the agent — add a filter_by_price tool alongside retrieve_docs and let the LLM compose them.
  3. In the agent's system prompt — give the LLM filter rules and trust it to apply them. Cheapest, least reliable.

Option 1 is the right answer for production; options 2 and 3 are useful for demos that need quick wins.

Source: SKILL.md on GitHub

skilld matched fixed text patterns in SKILL.md and file names. Patterns miss obfuscated code.

skilld run checks every file with the same patterns. It asks for approval before it loads a Skill with a behavior marked Needs approval.

No third-party reports yet.

Signed by skilld at ead8d7f. This ties the file your Agent reads to that commit on GitHub. It does not review the instructions.

Last checked against GitHub 3 days ago.

Activeupdated 2 months ago
metadata
{
  "author": "Google",
  "license": "Apache-2.0",
  "version": "0.2.0"
}

README badge

README badge for google/adk-recipes/product-search