The rules the reference agents enforce in code for third-party content, writes, grounding, identity, and memory, each with its module, plus two adversarial-eval rules. Load when handling untrusted tool results, guarding a write tool, or scoping what an agent remembers.
安装
npx skills add anthropics/commerce-agents --skill commerce-trust-safetynpx skills add anthropics/commerce-agents --skill commerce-trust-safety -a claude-code -gnpx skills add anthropics/commerce-agents --skill commerce-trust-safety -a codex -ggemini skills install https://github.com/anthropics/commerce-agents.git --path plugins/commerce-builder/skills/commerce-trust-safetygh skill install anthropics/commerce-agents commerce-trust-safetygit clone --depth 1 https://github.com/anthropics/commerce-agents /tmp/anthropics__commerce-agents
mkdir -p ~/.claude/skills ~/.agents/skills
cp -R /tmp/anthropics__commerce-agents/plugins/commerce-builder/skills/commerce-trust-safety ~/.claude/skills/commerce-trust-safety # Claude Code
cp -R /tmp/anthropics__commerce-agents/plugins/commerce-builder/skills/commerce-trust-safety ~/.agents/skills/commerce-trust-safety # Codex, Cursor, Gemini CLI, Copilot, OpenCode