Skill UI

A comprehensive framework for rigorously testing and benchmarking LLM agents. It moves beyond simple pass/fail checks to assess complex behaviors, reliability metrics, and capability consistency across multiple runs. Ideal for production monitoring and identifying subtle failure modes in advanced AI agents.

This skill defines structured behavioral frameworks (e.g., Brainstorm, Implement, Debug, Review, Teach, Ship) to guide the AI's output and approach based on the specific task phase. It optimizes performance by ensuring the AI adopts the correct mindset and output format—from initial ideation to final deployment—guaranteeing systematic, high-quality results across the entire development lifecycle.

Provides interpretable explanations for ML predictions using SHAP, LIME, and other techniques, helping debug model behavior, highlight feature importance, and communicate fairness insights to stakeholders.

Pyvene is a declarative framework designed for performing causal interventions on PyTorch neural networks. It allows researchers to conduct advanced experiments such as activation patching, causal tracing (ROME-style), and interchange intervention training (IIT). Use this library when you need to test causal hypotheses about model behavior, deeply interpret model components, or ensure reproducibility in advanced AI research.

This workflow guides users through building advanced AI voice agents using the Retell AI SDK. It demonstrates how to implement complex call flows, integrate function calling for tasks like booking appointments, and configure voice behavior for professional telephony applications.

SAELens provides a framework for training and analyzing Sparse Autoencoders (SAEs). SAEs decompose the dense, often polysemantic activations of large language models into sparse, monosemantic features. Use this when you need to discover the discrete, interpretable concepts a model has learned, study feature superposition, or analyze specific safety-relevant behaviors (like bias or deception) within deep neural networks.

TensorBoard delivers a unified interface to track training metrics, debug network behavior, compare experiments, inspect model graphs, and profile performance for PyTorch or TensorFlow workflows.

This tool provides a comprehensive framework for analyzing web server logs (Apache/Nginx) to detect various security threats. It uses advanced regex matching against OWASP attack signatures to identify SQL injection, Local File Inclusion, and Directory Traversal attempts, alongside behavioral analysis for brute-force attacks and web scanner fingerprints. The system also enriches data using GeoIP for source attribution, generating a prioritized report for SOC analysts.

This skill systematically probes an AI model's core behavioral patterns across six critical dimensions, including reasoning, refusal, formatting, and grounding. It generates a comprehensive, styled HTML report with radar charts and actionable insights, allowing users to understand a model's true capabilities, compare different models, or perform compliance and red-team assessments without needing an external API key.

This skill provides comprehensive tools to download, export, and deeply inspect existing Arize traces and spans. It is essential for understanding the full execution flow of complex LLM applications, debugging runtime errors, and performing root-cause analysis. Users can retrieve detailed data filtered by Trace ID, Span ID, or Session ID, enabling thorough analysis of conversational or multi-step agent behaviors.

This skill is designed to translate vague, colloquial, or descriptive biological details (such as behaviors, appearances, or cultural uses) into precise scientific or common species names. It is essential for constructing effective search queries in biological databases when the question does not use formal taxonomic names. By accurately identifying candidate species names from descriptive clues, users can drastically improve search relevance and accuracy.

This guide provides a comprehensive framework for debugging and analyzing AgentCore runtime sessions using CloudWatch Logs Insights. It details advanced techniques like OpenTelemetry (OTEL) span correlation, resolving session IDs to trace IDs, and structuring complex query queries. Use this when investigating agent execution flows, diagnosing tool call failures, analyzing latency, or building detailed behavioral timelines.