semantic-merge

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

• Third-party content exposure detected (high risk: 0.75). The required workflow reads repository state and commit messages/diffs/docs (including outsider-authored content from the incoming ref’s commits and any docs in that range) into the agent’s LLM context during Step 2 “Investigate intent on both sides” and Step 1 “Ground in repository state.”