edge-strategy-reviewer

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

• Third-party content exposure detected (high risk: 0.75). The skill loads and parses outsider-authored strategy draft YAML files from a user-supplied --drafts-dir/--draft path (load_draft_file() → yaml.safe_load()), then ingests their free-text fields (e.g., thesis, entry.conditions, validation_plan) into the review logic and outputs—so any malicious text in those YAMLs can be reflected into the LLM context downstream.