How does this skill handle document updates?

Since Gemini documents are immutable, the skill implements a delete-and-re-upload pattern to ensure your knowledge base stays current.

Are there costs associated with indexing?

Yes, indexing costs approximately $0.15 per 1M input tokens. This skill includes logic to estimate these costs before you upload large batches.

Which models are supported by this skill?

This skill requires Gemini 2.5 Pro or Gemini 2.5 Flash models; it is not compatible with Gemini 1.5 versions.

Does it handle storage quotas?

Yes, it accounts for the 3x storage multiplier (files + embeddings + metadata) to help you stay within your Gemini API tier limits.

What file formats can I index?

You can index over 100 formats, including PDF, Microsoft Word, Excel, and various source code files.

Google Gemini File Search

Name: Google Gemini File Search
Author: jezweb

byjezweb

•

117

•

Data Science & ML

Implements fully managed Retrieval-Augmented Generation (RAG) using Google Gemini for searchable document knowledge bases.

This skill enables developers to build sophisticated document Q&A systems and searchable knowledge bases using the Google Gemini File Search API. It provides a robust implementation of managed RAG with automatic chunking, embeddings, and citations across 100+ file formats including PDF, Word, and Excel. Beyond basic setup, it incorporates critical safeguards against common API errors such as document immutability, storage quota exceeding, and indexing cost surprises, ensuring efficient integration with frameworks like Next.js and Cloudflare Workers.

Key Features

01Built-in indexing cost estimation and operation polling logic.

02Supports 100+ file formats including PDFs, Word docs, Excel, and source code.

03Automated RAG workflow with managed chunking, embeddings, and semantic search.

04117 GitHub stars

05Comprehensive error prevention for document immutability and storage quotas.

06Optimized chunking strategies for technical, legal, and prose content.

Use Cases

01Building an internal AI-powered knowledge base from corporate documents.

02Creating a technical documentation Q&A bot for developer portals.

03Implementing semantic search within a Next.js or Cloudflare Workers application.

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add jezweb/claude-skills google-gemini-file-search

For use in Claude.ai and ChatGPT

Download Skill