plan-to-tickets

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

• Third-party content exposure detected (high risk: 0.85). SKILL.md requires reading the plan/spec files (step 1) and explicitly extracting their free-text task blocks to reconstruct the epic/ticket bodies that are then used by create-tickets.sh; those bodies are LLM-ingested content from repo files (which are outsider-authored unless the operating user authored them), so there is indirect prompt-injection exposure via plan/spec markdown text.