conductor-revert

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

• Third-party content exposure detected (medium risk: 0.65). The skill’s runtime workflow uses git log --oneline --grep=... to discover commits and then displays/uses commit messages and file paths; those commit messages are outsider-authored text (from other developers) that can be ingested into the agent’s LLM context via the generated “REVERT EXECUTION PLAN” and conflict/summary outputs.