Write and review code as if it gets pentested tomorrow. Load this whenever code touches a trust boundary: authentication, authorization, user input, database access, secrets or environment variables, file uploads, outbound requests (fetch/webhooks), API or webhook endpoints, or anything that runs raw SQL/shell/HTML.
安装
npx skills add kulchankas/paranoid --skill paranoidnpx skills add kulchankas/paranoid --skill paranoid -a claude-code -gnpx skills add kulchankas/paranoid --skill paranoid -a codex -ggemini skills install https://github.com/kulchankas/paranoid.git --path skills/paranoidgh skill install kulchankas/paranoid paranoidgit clone --depth 1 https://github.com/kulchankas/paranoid /tmp/kulchankas__paranoid
mkdir -p ~/.claude/skills ~/.agents/skills
cp -R /tmp/kulchankas__paranoid/skills/paranoid ~/.claude/skills/paranoid # Claude Code
cp -R /tmp/kulchankas__paranoid/skills/paranoid ~/.agents/skills/paranoid # Codex, Cursor, Gemini CLI, Copilot, OpenCode