benchmark-datasets
Standard datasets and benchmarks for evaluating AI security, robustness, and safety
npx skills add majiayu000/claude-skill-registry --skill benchmark-datasets --agent claude-code
Same command for any agent — swap --agent for codex, cursor, copilot.
Weekly change comes from our own snapshots, not the repository page — it measures attention, not adoption.
# AI Security Benchmark Datasets Use **standardized benchmarks** to evaluate and compare AI system security, robustness, and safety. ## Quick Reference ```yaml Skill: benchmark-datasets Agent: 04-evaluation-analyst OWASP: LLM01 (Injection), LLM02 (Disclosure), LLM04 (Poisoning), LLM05 (Output), LLM09 (Misinfo) NIST: Measure Use Case: Standardized security evaluation ``` ## Benchmark Taxonomy ``` [AI Security Benchmarks] │ ┌───────────────────┼───────────────────┐ ▼ ▼ ▼ [Safety] [Robustness] [Privacy] │ │ │ ┌────┴────┐ ┌────┴────┐ ┌────┴────┐ ▼ ▼ ▼ ▼ ▼ ▼ [Toxicity] [Bias] [Adversarial] [OOD] [MIA] [Extraction] │ ┌───────┴───────┐ ▼ ▼ [Jailbreak] [Injection] ``` ## Safety Benchmarks ### LLM Safety Evaluation ```yaml HarmBench: description: "Comprehensive harmful behavior evaluation" size: 400+ behaviors categories: - Chemical/Biological weapons - Cyberattacks - Harassment - Fraud/Deception metrics: - Attack Success Rate (ASR) - Defense Success Rate (DSR) url: https://harmbench.org owasp: [LLM05] ToxiGen: description: "Machine-generated toxic content detection" size: 274k examples categories: - Identity attacks - Hate speech - Threats metrics: - Toxicity Score - False Positive Rate ow
- Quick Reference
- Benchmark Taxonomy
- Safety Benchmarks
- LLM Safety Evaluation
- Benchmark Framework
- Jailbreak Benchmarks
- Adversarial Robustness Benchmarks
- Privacy Benchmarks
- Evaluation Dashboard
- Custom Benchmark Creation
- Troubleshooting
- Integration Points
What does the benchmark-datasets skill do?
Standard datasets and benchmarks for evaluating AI security, robustness, and safety
How do I install it?
Run `npx skills add majiayu000/claude-skill-registry --skill benchmark-datasets --agent claude-code` — it drops the skill into your project so the agent can pick it up. Swap the --agent value for codex, cursor or copilot if you use one of those.
Where does this skill come from?
From majiayu000/claude-skill-registry, a repository with 534 stars. We read it straight from the repository tree rather than a submitted listing, so what you see here is what is actually published.
Is a popular skill a good skill?
Not necessarily. Stars measure attention, not adoption — a repository can trend for a week and be abandoned. That is why we show the weekly change from our own snapshots next to the total, instead of a single flattering number.