spree-testing

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

• Third-party content exposure detected (high risk: 0.85). High likelihood: the workflow’s “Fetch live docs” step downloads public web content from https://spreecommerce.org/docs/developer/tutorial/testing at runtime, and that fetched page text can be ingested into the agent’s LLM context (public web content → indirect prompt injection risk).

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

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly instructs the agent to "Fetch https://spreecommerce.org/docs/developer/tutorial/testing" as a live docs dependency to guide testing instructions, so remote content would be fetched at runtime and could directly control the agent's behavior.