{
  "id": "cyberseceval",
  "name": "CyberSecEval",
  "category": "evals-observability",
  "summary": "Meta's benchmark suite for LLM cybersecurity risk and defensive skill: attack compliance, insecure code, prompt injection and SOC tasks.",
  "description": "CyberSecEval 4 builds on versions 1-3 and adds defensive benchmarks such as CyberSOCEval. Tests include MITRE ATT&CK compliance and false-refusal rates, instruct and autocomplete secure-code tests, and prompt-injection tests. Code and benchmarks are MIT, in PurpleLlama/CybersecurityBenchmarks.",
  "url": "https://meta-llama.github.io/PurpleLlama/CyberSecEval/",
  "repo": "https://github.com/meta-llama/PurpleLlama",
  "docs": "https://github.com/meta-llama/PurpleLlama/tree/main/CybersecurityBenchmarks",
  "tags": [
    "benchmark",
    "security",
    "prompt-injection",
    "secure-code",
    "meta"
  ],
  "license": "MIT",
  "pricing": "open-source",
  "status": "active",
  "related": [
    "llamafirewall",
    "garak",
    "pyrit",
    "agentdojo"
  ],
  "sources": [
    "https://github.com/meta-llama/PurpleLlama/tree/main/CybersecurityBenchmarks",
    "https://github.com/meta-llama/PurpleLlama/blob/main/CybersecurityBenchmarks/LICENSE",
    "https://meta-llama.github.io/PurpleLlama/CyberSecEval/"
  ],
  "added": "2026-11-02",
  "updated": "2026-11-02",
  "last_verified": "2026-11-02",
  "submitted_by": "agentica-curator",
  "maintainer": "Meta",
  "longform": [],
  "links": {
    "html": "https://indexagentica.com/entries/cyberseceval/",
    "markdown": "https://indexagentica.com/entries/cyberseceval.md",
    "json": "https://indexagentica.com/api/entries/cyberseceval.json",
    "source": "https://github.com/Drudley/indexagentica/blob/main/content/evals-observability/cyberseceval.json"
  }
}