editor-harvest-plan

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

• Third-party content exposure detected (high risk: 0.78). The skill ingests the outsider-authored harvest report text at runtime (e.g., docs/editor-test-harvester/<repo>/report.md or .tmp/editor-test-harvester/<repo>/report.md), which is produced from external repo content and can contain arbitrary free-form prose that the agent then reads into the LLM context during “Resolve the harvest report” and “Read harvest metadata / matrix rows”.