{
  "name": "llm-dark-patterns",
  "description": "Out-of-band Claude Code Stop / SubagentStop hook judges that block 30 LLM dark patterns at the closeout boundary: false-success without evidence, sycophancy, paternalism, permission-loops, vibe time estimates, fake recall, fake stats, fake citations, count-vs-enumeration self-consistency drift, context loss after compaction, multi-agent rollup failures, power-user polish failures, and unreachable new symbols (advisory). Deterministic verdicts, no network calls, no model in the verdict path, Apache-2.0, paper-grade claim ledger.",
  "version": "1.2.1",
  "author": {
    "name": "waitdeadai",
    "url": "https://github.com/waitdeadai"
  },
  "homepage": "https://github.com/waitdeadai/llm-dark-patterns",
  "repository": "https://github.com/waitdeadai/llm-dark-patterns",
  "license": "Apache-2.0",
  "category": "security",
  "tags": [
    "safety",
    "governance",
    "hooks",
    "stop-hooks",
    "ai-safety"
  ],
  "keywords": [
    "claude-code",
    "hooks",
    "dark-patterns",
    "anti-sycophancy",
    "anti-hallucination",
    "verification",
    "governance",
    "ai-safety",
    "llm-safety",
    "abstention"
  ]
}