by affaan-m
healthcare-eval-harness is a patient safety evaluation harness for healthcare app deployments. It helps teams verify CDSS accuracy, PHI exposure, data integrity, clinical workflow behavior, and integration compliance before release. Critical failures block deployment, making it useful for healthcare-eval-harness for Model Evaluation and CI safety gates.
by affaan-m
The eval-harness skill is a formal evaluation framework for Claude Code sessions and eval-driven development. It helps you define pass/fail criteria, build capability and regression evals, and measure agent reliability before shipping prompt or workflow changes.
by affaan-m
agent-eval is a skill for benchmarking coding agents head-to-head on reproducible tasks, comparing pass rate, cost, time, and consistency. Use the agent-eval skill to evaluate Claude Code, Aider, Codex, or another agent in your own repo with clearer evidence than ad hoc prompting.
by wshobson
The evaluation-methodology skill explains PluginEval scoring for Model Evaluation, including layers, rubrics, composite scoring, badge thresholds, and practical guidance for interpreting results and improving weak dimensions.
by wshobson
Use the llm-evaluation skill to design repeatable evaluation plans for LLM apps, prompts, RAG systems, and model changes with metrics, human review, benchmarking, and regression checks.
by github
ai-prompt-engineering-safety-review is a prompt audit skill for reviewing LLM prompts for safety, bias, security weaknesses, and output quality before production, evaluation, or customer-facing use.
by github
agentic-eval is a GitHub Copilot skill that shows how to build evaluation loops for AI outputs using reflection, rubric-based critique, and evaluator-optimizer patterns.
by googleworkspace
gws-modelarmor helps you work with Google Model Armor in the googleworkspace/cli ecosystem. Use it to sanitize prompts, sanitize model responses, and create templates with less guesswork than a generic prompt. It is designed for repeatable, policy-aware usage and Security Audit workflows.
by huggingface
huggingface-community-evals helps you run Hugging Face Hub model evaluations locally with inspect-ai or lighteval. Use it for backend selection, smoke tests, and a practical guide to vLLM, Transformers, or accelerate. Not for HF Jobs orchestration, model-card PRs, .eval_results publishing, or community-evals automation.
by huggingface
The huggingface-best skill helps you find the best model for a task by checking Hugging Face benchmark leaderboards and filtering by device limits and model size. Use it for model recommendations in coding, reasoning, chat, OCR, RAG, speech, vision, or multimodal work when you need a practical shortlist, not a generic model list.
by mukul975
analyzing-campaign-attribution-evidence helps analysts weigh infrastructure overlap, ATT&CK consistency, malware similarity, timing, and language artifacts for defensible campaign attribution. Use this analyzing-campaign-attribution-evidence guide for CTI, incident analysis, and Security Audit reviews.
by trailofbits
The libafl skill helps you plan and build modular fuzzers with LibAFL for custom targets, mutation strategies, and security audit workflows. Use this libafl guide to move from target details to a practical harness, feedback model, and run plan with fewer assumptions.
by NeoLabHQ
judge-with-debate evaluates solutions through structured multi-agent debate, using a shared specification, evidence-based counterarguments, and up to 3 rounds to reach consensus. It is well suited for code review, rubric-based assessment, and judge-with-debate for Multi-Agent Systems workflows.
by muratcankoylan
The evaluation skill helps you design and run agent evaluations for non-deterministic systems. Use it for evaluation install planning, rubrics, regression checks, quality gates, and evaluation for Skill Testing. It fits LLM-as-judge workflows, multi-dimensional scoring, and practical evaluation usage when you need repeatable results.
by mukul975
detecting-ai-model-prompt-injection-attacks is a cybersecurity skill for screening untrusted text before it reaches an LLM. It uses layered regex, heuristic scoring, and DeBERTa-based classification to flag direct and indirect prompt injection attacks. Useful for chatbot input validation, document ingestion, and Threat Modeling.
by wshobson
ml-pipeline-workflow is a practical guide to designing end-to-end MLOps pipelines for data prep, training, validation, deployment, and monitoring, with orchestration patterns for repeatable workflow automation.
