chezmoi-management

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.90). The SKILL.md explicitly clones a public GitHub repo ("chezmoi init https://github.com/mizchi/chezmoi-dotfiles.git --apply") and runs run_after_apm-install.sh which invokes "apm install --global --target claude" (and dot_apm/apm.yml lists external APM packages), so the agent will fetch and load untrusted public third-party skills that can influence its behavior.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly runs "chezmoi init https://github.com/mizchi/chezmoi-dotfiles.git --apply" (cloning that repo at runtime) and the repo's post-apply hook runs "apm install --global --target claude", which will fetch and install external APM agent-skills (remote code and prompt/instruction content) into the agent runtime, so the URL https://github.com/mizchi/chezmoi-dotfiles.git is a runtime external dependency that can control prompts/execute code.