grill-with-docs-codex

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

• Third-party content exposure detected (medium risk: 0.30). At runtime, Act 2’s review prompt instructs Codex to read the plan and repository files (including CONTEXT.md/ADRs), while Act 1 explicitly updates/uses CONTEXT.md/ADRs based on the user-provided plan text, so the workflow ingests outsider-authored free text via the user interaction and then feeds that derived text into the LLM review loop.