auto-loom-proof

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

• Third-party content exposure detected (medium risk: 0.65). SKILL.md points to using outsider-authored free text in factory/HANDOFF.md and factory/REVIEW.md, and it instructs the agent to read the “contract” to select criteria and then narrate/paraphrase them, so that text can be injected into the LLM’s runtime context via those input markdown files.