claude-skills-troubleshooting

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

• Third-party content exposure detected (high risk: 1.00). This skill loads plugin metadata and SKILL.md files from public marketplace git clones (e.g., GitHub repos referenced in references/architecture.md and known_marketplaces.json) and explicitly adds that SKILL.md content to the system prompt/context for use in responses, so untrusted third-party skill content can directly influence agent behavior.