agent-skills-context-engineering

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

• Third-party content exposure detected (high risk: 0.90). SKILL.md explicitly instructs fetching skill files from a public raw GitHub URL (curl https://raw.githubusercontent.com/...), and the included SkillRegistry.load_skill pattern reads those files into the agent at runtime, meaning arbitrary user-generated third-party content can be ingested and materially influence agent behavior.