scan-new-specs

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

• Third-party content exposure detected (high risk: 0.75). Outsider-authored free text from public/third-party sources (merged PRs in warpdotdev/warp and warp-server, specifically specs/<id>/PRODUCT.md content) is read at runtime and then used to generate Slack messages and to drive the write-feature-docs ambient drafting workflow, so it can enter the agent/LLM context indirectly via the spec body.