MCP App StoreMCP App StoreThis skill provides a structured, interactive workflow for creating experiment.yaml configuration files for the AEC Bench tool. It guides the user through the process of selecting tasks, choosing AI agents and models, and defining execution settings to ensure a valid benchmark run.
Use this skill when a user wants to start a new benchmark experiment, modify an existing configuration, or plan a trial run with a dry run preview.
aec-bench.toml), task selection via datasets or disk scanning, agent configuration (including model selection from a compatibility matrix), and execution settings.Designed for agents capable of executing shell commands and reading/writing YAML files, such as Claude Code or similar autonomous IDE agents.
This skill has not been reviewed by our automated audit pipeline yet.
LLM Evaluation
Evaluation framework and tools for systematically measuring LLM performance using automated metrics, human judgment, and A/B testing.
ROCm Triton Kernels (RMSNorm, RoPE 3D, GEGLU, AdaLN)
Guidance and examples for writing, benchmarking, and integrating optimized Triton kernels on ROCm (MI355X, R9700) for diffusers and transformers workloads.
Skill Creator
Create, improve, and evaluate Agent Skills with a guided workflow: capture intent, draft SKILL.md, run evals and benchmarks, and optimize triggering description
Performance Optimizer
Guides profiling and targeted optimizations for code and systems — measure, identify bottlenecks, and verify improvements across Python, Node, shell, and system
Hugging Face Evaluation Manager
Extract, import, and add structured model evaluation results to Hugging Face model cards; run or import benchmark evaluations and generate model-index YAML for
LLM Council
Run parallel queries across multiple LLMs with a live dashboard to compare outputs, synthesize consensus, and perform anonymous model voting.
TAO Performance Audit
Structured performance-audit methodology: measure, identify bottlenecks, optimize the true hotspot, and verify improvements with benchmarks.
Run Benchmarks
Launch, manage, and rerun CodeScaleBench benchmark suites with safety guardrails, paired baseline+full execution, and orchestration utilities.