agent-platform-deploy

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The prompt explicitly shows inserting a Hugging Face access token as a literal CLI argument (--hugging-face-access-token="YOUR_HF_TOKEN") and includes examples that embed an access token in an Authorization header, which requires the LLM/agent to handle or output secret values verbatim.

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

• Third-party content exposure detected (medium risk: 0.65). The required runtime workflow fetches and ingests free-form text from an outsider-authored source via curl calls to ${ENDPOINT}/ui/${MODEL} / ${ENDPOINT}/ui/${MODEL_COPY} (and related UI endpoints) to verify existence and describe results, which can include arbitrary page/body text not authored by the operating user.