Do I need to manually normalize Gemini embeddings?

Normalization is required when using output dimensions other than the default 3072 to ensure accurate cosine similarity calculations.

What are 'Task Types' in Gemini embeddings?

Task types are parameters like RETRIEVAL_QUERY or CLASSIFICATION that optimize the embedding vector for your specific application use case.

How does this handle the Gemini API rate limits?

The skill provides implementation patterns for chunking requests and using exponential backoff to navigate Free, Tier 1, and Tier 2 API quotas.

Which Gemini embedding model does this skill support?

This skill specifically supports the gemini-embedding-001 model, which is the stable, production-ready version for generating high-quality text vectors.

Can I use these embeddings with Cloudflare Vectorize?

Yes, it includes specific guidance and code patterns for integrating Gemini embeddings with Cloudflare Workers and Vectorize storage.

Google Gemini Embeddings

Name: Google Gemini Embeddings
Author: brendadeeznuts1111

bybrendadeeznuts1111

0•

Data Science & ML

Builds production-ready RAG systems and semantic search using optimized Gemini embedding-001 models and vector storage patterns.

This skill provides a comprehensive implementation framework for integrating Google's gemini-embedding-001 model into Claude Code workflows. It streamlines the creation of Retrieval-Augmented Generation (RAG) systems by managing complex tasks like Matryoshka Representation Learning, batch processing, and vector normalization. Designed for high-reliability production environments, it includes specific workarounds for documented API bugs, such as batch ordering corruption and memory limits, while providing native patterns for Cloudflare Vectorize integration.

Key Features

01Pre-configured error handling for 13 documented SDK and API edge cases

020 GitHub stars

03Support for 8 specialized task types including Retrieval Query and Document

04Cloudflare Workers and Vectorize integration patterns for edge deployments

05Robust batch processing logic with exponential backoff and rate limit handling

06Flexible dimensionality (768-3072) via Matryoshka Representation Learning

Use Cases

01Implementing semantic search, document clustering, and duplicate detection

02Developing high-accuracy Retrieval-Augmented Generation (RAG) applications

03Scaling high-volume text embedding pipelines for large-scale data analysis

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add brendadeeznuts1111/tier-1380-omega google-gemini-embeddings

For use in Claude.ai and ChatGPT

Download Skill