remix-ad

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

• Third-party content exposure detected (high risk: 0.75). The required workflow fetches the source sample’s remix payload from the Goose Ads library via scripts/fetch_sample.py → scripts/lib.py (api_get("/api/ads-library/samples")), and that payload includes outsider-authored free text like extracted_script/how_to which is then read into the planner’s JSON and used to generate the new ad.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). The skill's runtime calls to the Gooseworks API (base "https://app.gooseworks.ai" and the list endpoint "/api/ads-library/samples") fetch remix payloads that include recipe/remix_spec.skills which directly drive which atom-skills and production templates the agent will execute, so remote content controls the agent's instructions at runtime.