{
  "name": "hs",
  "version": "1.4.9",
  "description": "Pre-execution safety layer that blocks dangerous shell commands and credential file reads using pattern matching + LLM analysis. Fail-closed design.",
  "author": {
    "name": "Francesco Marinoni Moretto",
    "email": "contact@clarity-gate.org"
  },
  "homepage": "https://github.com/frmoretto/hardstop",
  "repository": "https://github.com/frmoretto/hardstop",
  "license": "CC-BY-4.0",
  "keywords": [
    "safety",
    "security",
    "shell",
    "commands",
    "protection",
    "guardrails",
    "pre-execution",
    "fail-closed"
  ]
}