Skill UI

Optimize Perplexity Sonar API calls by selecting appropriate models, caching query hashes, streaming long research, and batching deduplicated requests for faster, cost-effective research or fact-checking workflows.

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.

Covers prompt and response caching plus Cache Augmented Generation patterns for LLMs such as Claude and OpenAI to reuse stable context, reduce repeat work, and cut latency/cost on recurring queries.

Meta’s Prompt Guard is an 86M-parameter classifier that spots prompt injections and jailbreak attempts before they reach LLMs, filtering user input, third-party data, or RAG documents with multilingual support and sub-2ms latency for fast deployments.

Guide for optimizing Retell AI voice agents by tuning LLM prompts, agent settings, WebSocket pooling, and caching/analytics so API integrations stay responsive under load.

Comprehensive skill covering the full lifecycle of visual AI systems. Expertise includes object detection (YOLO, Faster R-CNN), instance and semantic segmentation (SAM, Mask R-CNN), and video analysis. The workflow guides users through dataset preparation, selecting optimal architectures, training using frameworks like PyTorch, and achieving production readiness via model optimization (ONNX, TensorRT) and deployment on various edge or cloud targets.

SGLang is a high-performance serving framework for LLMs and VLMs that uses RadixAttention prefix caching to deliver structured outputs (JSON/regex/grammar) and agentic workflows with function calls, providing up to 5× faster inference than vLLM across multi-GPU production deployments.

Automates Textrazor tasks through Composio’s Rube MCP pipeline, using RUBE_SEARCH_TOOLS, RUBE_MANAGE_CONNECTIONS, and RUBE_MULTI_EXECUTE_TOOL so each workflow picks up the current schema and active toolkit connection.

TransformerLens is the standard library for performing mechanistic interpretability research on large language models. It provides clean interfaces to inspect and manipulate transformer internals—such as attention patterns, residual streams, and MLP outputs—via activation caching and HookPoints. This tool is essential for reverse-engineering model algorithms, performing causal tracing, and studying the internal circuits responsible for model behavior.

Guides Claude to automatically tune machine learning model hyperparameters via grid, random, or Bayesian search, generating runnable Python code, executing validations, and reporting best configurations plus performance metrics for accuracy or other goals.

Guide to optimizing TwinMind transcription, balancing accuracy and speed through audio preprocessing, scenario-specific model configs, streaming tweaks, and caching to cut redundant work.

DeepChem is a Python library for molecular machine learning, providing featurizers, dataset loaders, splitters, and models spanning traditional ML to GNNs for property prediction, drug discovery, and materials design.