Skill UI

Automate OpenAI API workflows through the Composio MCP integration, covering text and multimodal responses with structured outputs, embeddings, image generation, and model discovery.

A comprehensive workflow designed for clinical professionals to search, retrieve, and synthesize medical evidence. It allows users to query clinical literature with specialized filters (e.g., specialty, evidence level), check potential drug interactions against patient profiles, and look up the latest specialty guidelines from major health organizations (like ACC/AHA, NICE). The output provides structured citations, confidence scores, and evidence grading to support informed clinical decision-making.

High-performance RLHF framework that combines Ray, vLLM, and ZeRO-3 to accelerate PPO/GRPO/RLOO/DPO training for 7B-70B+ models on distributed GPU clusters, providing a single-stack workflow for reward models and policy optimization.

This skill provides advanced strategies for managing large language model (LLM) context windows and token usage via OpenRouter. Functions include context window lookup, context-aware model selection, conversation trimming, document chunking for RAG, and prompt caching. Use this when dealing with long conversations or large documents to maximize quality while minimizing API costs and preventing context limit overflow.

Implements standardized function/tool calling across various large language models (LLMs) via OpenRouter. This unified approach allows developers to build sophisticated, multi-step agents and complex workflows, regardless of the underlying provider (e.g., GPT-4o, Claude 3.5). It simplifies the process of achieving structured output and tool-augmented reasoning in production systems.

Enables Claude to orchestrate multi-agent systems using AI SDK v5, setting up agents, handoffs, intelligent routing, and workflows across OpenAI, Anthropic, and Google when users request complex agent coordination.

Covers HuggingFace PEFT workflows for LoRA/QLoRA-based LLM fine-tuning, enabling sub-1% parameter updates on limited GPUs plus adapter serving tips.

This workflow executes a single-query search using the Perplexity Sonar model, ensuring all returned answers are web-grounded and accompanied by inline citations. It is essential for building high-trust applications, fact-checking tools, or any system requiring verifiable, real-time information retrieval. The process includes API interaction, detailed response parsing (answer, citations, search results), and formatting the output into a usable, source-linked format.

A comprehensive toolkit designed to manage the complex data flow associated with Perplexity search workflows. This skill implements critical security measures, including PII sanitization in user queries, validates external citations for accuracy, optimizes performance via result caching with freshness policies, and manages conversation context to maintain coherence and reliability. Essential for building robust, compliance-aware AI applications.

This skill provides production-ready patterns for integrating with the Perplexity Sonar API in both TypeScript and Python. It covers best practices for creating robust wrappers around the standard OpenAI client, including dedicated handling for citations, search results, and related questions. It also includes essential utilities like implementing exponential backoff for reliable retries. Use this when setting up complex search-augmented generation workflows or standardizing team coding practices around Perplexity usage.

Prediction Monitor Assistant automates machine learning deployment monitoring tasks, offering step-by-step guidance, best-practice patterns, production-ready code and configuration generation, and validation to keep serving and inference workflows reliable.

Document describes structured prompts, chain-of-thought guidance, few-shot examples, system prompt design, and tooling advice for building reliable prompt engineering workflows in LLM-augmented products.