Skill UI

Sentence Transformers provides 5000+ pre-trained models that generate sentence, text, and image embeddings for RAG, semantic search, clustering, and multilingual production workloads without relying on external APIs.

Vexor is a powerful command-line interface designed for advanced semantic search across local files. By utilizing vector embeddings and integrating with large language models (like Claude/Codex), it allows users to search files based on meaning and context, rather than just exact keywords. This is ideal for large codebases or document repositories where traditional keyword search methods are insufficient.

Combines ESM3 and ESM C protein language models to generate sequences, predict structures, condition functions, and extract embeddings via local or Forge API inference for protein design and engineering workflows.

Geniml builds machine learning workflows around BED and scATAC datasets to learn embeddings for regions, metadata, and cells, enabling similarity searches, clustering, consensus peaks, and downstream analysis of genomic intervals.

Comprehensive guide to LLM Operations (LLMOps). Covers advanced topics like Retrieval-Augmented Generation (RAG), vector database management (Pinecone, Chroma), embeddings, fine-tuning, prompt engineering, and cost optimization. Essential for bridging the gap between AI prototypes and reliable, scalable, production-grade AI systems.

ESM provides advanced protein language models (ESM3, ESMC) for deep biological understanding, generation, and structural prediction. This skill enables users to generate novel protein sequences with specific functions, perform structure prediction from sequence, execute inverse folding (sequence design from structure), and extract high-quality protein embeddings for downstream machine learning tasks. Ideal for drug discovery and advanced bioinformatics research.

A comprehensive guide providing minimal working examples for core Cohere API v2 endpoints, including Chat completion, text embedding, and search reranking. This resource details basic setup, usage patterns, and includes runnable code examples in both TypeScript and Python, making it perfect for new AI integrations or learning advanced LLM functionality.

This guide demonstrates how to build advanced AI agents using Anthropic's Tool Use (Function Calling) feature. It allows Claude to dynamically determine when external actions are needed, such as querying databases, calling APIs, or reading files. The pattern involves sending the user query, receiving a `tool_use` request, executing the corresponding function in your code, and submitting the result back to Claude for a final, coherent response. Essential for building multi-step, agentic workflows.

This comprehensive workflow guides the implementation of a complete Retrieval-Augmented Generation (RAG) pipeline using Cohere's core services. It demonstrates the full process: embedding documents, searching for relevant context, reranking results for precision, and finally generating a grounded answer with precise citations using the Chat API. Ideal for building knowledge-intensive, enterprise-grade Q&A systems.

Provides strategies to optimize costs when utilizing the Cohere API. This guide covers model tier selection (economy, standard, premium), implementing token budgets, reducing embedding storage costs (e.g., using int8), and establishing comprehensive usage trackers. It is essential for maintaining cost predictability in production AI deployments.

This comprehensive guide provides a deep dive into migrating AI applications from major LLM providers (like OpenAI or Anthropic) to the Cohere platform. It covers critical technical steps, including API call adaptation, re-vectorizing embeddings, migrating tool definitions, and implementing the Adapter Pattern for seamless, gradual re-platforming of existing AI services.

A comprehensive guide for optimizing Cohere API v2 performance. Learn advanced strategies to improve throughput, reduce latency, and manage costs when using Chat, Embed, and Rerank endpoints. Techniques covered include dynamic model selection based on latency budgets, implementing streaming for faster perceived response times, efficient batch processing for embeddings, vector compression, and robust caching mechanisms.