apple-app-store-reviews

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.80). The skill explicitly allows and demonstrates passing the FREVANA_TOKEN as a --token command-line argument (including an example with "your bearer token"), which would require the LLM to include secret values verbatim in generated commands/outputs (high exfiltration risk), even though it also suggests reading the token from the environment and cautions not to echo it.

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

• Third-party content exposure detected (medium risk: 0.30). The runtime workflow posts a user-specified product_id to https://ai-factory.frevana.com/service/apple-reviews and then ingests the returned app review JSON from that API response, which originates from Apple user-submitted review text.