Skip to content

strato-space/fast-agent-vertex-rag

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

8 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

fast-agent-vertex-rag

Minimal AgentCard repo that wraps Google Vertex RAG as a function tool for fast-agent.

This repo follows the AgentCard RFC: https://github.com/evalstate/fast-agent/blob/main/plan/agent-card-rfc.md

Setup

  1. Install dependencies:
uv sync
  1. Configure Vertex AI settings:
cp fastagent.secrets.yaml.example fastagent.secrets.yaml

Update fastagent.secrets.yaml with your Google Cloud project and region. Authenticate with Google (for example, gcloud auth application-default login) and ensure the Drive folder is accessible by your credentials. Enable Vertex AI RAG + Google Drive APIs and required roles in your project. If access is denied, share the Drive folder with the Google service account from the error message.

Run

Batch

uv run fast-agent go --card agents --message "Produce a short top 5 prioritized list about customer pain points. From RAG, select 50 relevant chunks about customer pain points. Deduplicate. Links: [name](<link>). Compact output. Drive ID: 1J3ubtdkmFuWDjfW3_qT2Fhsdn2pbtv-8."

Interactive (with Read-Eval-Print-Loop)

uv run fast-agent go --card agents --watch

Video: AgentCard: fast-agent-vertex-rag sample

AgentCard: fast-agent-vertex-rag sample

Design (why this is minimal)

  • agents/vertex-rag.md is the AgentCard.
  • function_tools loads agents/vertex_rag_tool.py:mini_rag via the function tool loader.
  • pyproject.toml declares all runtime dependencies, including fast-agent-mcp.
  • fastagent.secrets.yaml supplies Vertex AI project/region for the tool.

References

AgentCard

fast-agent.ai

Google Vertex RAG

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages