agent-skills-architect

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

• Third-party content exposure detected (high risk: 0.80). The SKILL.md explicitly directs the agent to "Start with the official Agent Skills LLM docs at https://agentskills.io/llms-full.txt" and to "resolve the linked repository and inspect the actual skill directory" (skills.sh handling and Operating stance/workflows.md), which requires fetching and interpreting external public websites and repositories (untrusted third‑party content) that can materially influence subsequent actions.