Skill UI

Covers designing agent tools from schema creation through validation and error handling, emphasizing descriptions that keep LLMs accurate and reusable via MCP standards.

A comprehensive toolkit for evaluating the performance of code generation models. It supports benchmarking against industry standards like HumanEval, MBPP, and MultiPL-E across multiple languages. Use this tool to quantitatively compare different LLM coding abilities and measure code synthesis quality using pass@k metrics.

Provides automated guidance for configuring learning rate schedulers during ML training, covering best practices, setup patterns, and framework-ready code snippets for PyTorch, TensorFlow, and Scikit-learn while validating against common standards.

Automates guidance for model export helper tasks across ML deployment, delivering step-by-step best practices, production-ready code/configuration, and validation against common standards.

Assists with torchscript exporter tasks for ML deployment, offering automated guidance, code/config generation, and validation against industry standards whenever torchscript exporters are mentioned.

This skill provides comprehensive tools for documenting and managing model performance on Hugging Face. It supports multiple methods for adding structured evaluation data, including extracting tables from README content, importing benchmark scores via APIs (like Artificial Analysis), and executing high-performance custom evaluations using advanced backends such as vLLM or lighteval. It ensures that all metadata is structured and compatible with model leaderboards and industry standards.

This skill provides a comprehensive framework for optimizing, improving, and debugging Large Language Model (LLM) prompts. It leverages production trace data, system evaluations, and human annotations to extract prompts from spans. Users can run a data-driven optimization loop, ensuring the AI responses meet desired quality standards. Ideal for prompt tuning, system prompt refinement, and improving overall AI output quality.

This workflow converts complex clinical encounter notes into plain-language, patient-facing After-Visit Summaries (AVS). It structures key information such as diagnosis, next steps, and medications. The system supports multi-language translation and facilitates pushing the final, readable summary directly to patient portals using FHIR standards.

Provides a comprehensive architectural blueprint for integrating ambient AI solutions (Abridge) into multi-site health systems. It guides the full process from ambient audio capture to structured note generation, covering data flow, component design, FHIR R4 standards, HIPAA compliance, and integration patterns with major EHRs like Epic and Athena.

This pattern library provides production-ready, structured patterns for integrating with the Abridge clinical AI API. It establishes type-safe client wrappers, implements mandatory HIPAA-compliant audit logging (ensuring no Protected Health Information, or PHI, is logged), and incorporates robust error handling with exponential backoff retries. Use this when building secure, scalable, and standards-compliant healthcare AI applications.

This tool generates production-grade, spec-compliant agent definition files (.md) adhering to the Anthropic 2026 schema. It supports both the creation of new custom subagents and the validation of existing agent files against marketplace compliance rules. Use it when building complex agent architectures, reviewing agent quality, or ensuring agents meet high industry standards.

This comprehensive checklist guides the safe and robust deployment of Cohere API v2 integrations from staging to production. It covers critical steps including API key management, code quality standards, performance tuning, implementing health checks and circuit breakers, and defining full rollback procedures, ensuring mission-critical AI features are reliable for launch.