om-app-spec-writing

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

• Third-party content exposure detected (high risk: 0.70). The required workflow reads repo-local and tracker content (issue/PR titles/bodies/comments, README/agent docs, config, CI logs) as “untrusted content boundary” and then passes the completed App Spec plus repo agent docs into dispatched subagents, so outsider-authored free text from those sources can enter the LLM context via the repo/tracker ingestion path.