agent-platform-rag-engine-management
Manage and query Agent Platform RAG Engine Corpora
by googleFree
Description
Agent Platform RAG engine management is an agent skill for working with retrieval-augmented generation corpora on Google Cloud Agent Platform. It lists and inspects RAG corpora and their files, and it runs queries that retrieve grounded contexts for a question. The skill requires the Agent Platform Python SDK, the vertexai package, for every RAG Engine operation, rather than raw REST calls or MCP tools, because it is meant to run in external clients. Its safety tiers separate operations by risk. Read-only actions such as listing corpora and files, getting a corpus and running a retrieval query run immediately. Operations that are read-only but use compute, such as generating content with the model client, are a separate tier that consumes resources. The skill is useful for developers who maintain a knowledge base for an assistant and want an agent to inspect it, check what has been indexed, or test retrieval before wiring it into an application. It is published by Google on skills.sh under the google skills repository, with about 8.2K installs. The listing shows Gen Agent Trust Hub, Socket and Snyk all passing.…
Highlights
- Lists and inspects RAG corpora and files
- Retrieves grounded contexts for queries
- Uses the vertexai Python SDK
- Safety tiers for reads, compute and changes
- Good for testing retrieval before deployment
Ratings & Reviews
No ratings yet
Be the first to rate agent-platform-rag-engine-management.
People also visit
embedding-strategies
Select and optimize embedding models for semantic search and RAG applications
hybrid-search-implementation
Combine vector and keyword search for improved retrieval
agent-platform-eval-flywheel
Measures and improves the quality of AI models and agents on Google Cloud
agent-platform-rag-engine-management FAQ
What does it manage?
RAG Engine corpora and the files inside them, and retrieval queries that return grounded contexts.
Which SDK does it use?
The Agent Platform Python SDK, the vertexai package, which the skill requires instead of raw REST calls or MCP tools.
Does a read need confirmation?
Read-only actions such as listing and retrieval run without confirmation. Actions that use compute or change data are in stricter tiers.
How do I install it?
Run npx skills add https://github.com/google/skills --skill agent-platform-rag-engine-management.
What security audits has it passed?
skills.sh lists three audits: Gen Agent Trust Hub Pass, Socket Pass and Snyk Pass.
What are the best agent-platform-rag-engine-management alternatives?
Popular alternatives to agent-platform-rag-engine-management include ai-engineer, embedding-strategies and hybrid-search-implementation. See all agent-platform-rag-engine-management alternatives