competitor-product-monitor

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

• Third-party content exposure detected (high risk: 1.00). The skill's required workflow Step 3 instructs the agent to run TinyFish against competitors' public websites (e.g., their homepage, /blog, /changelog, /newsroom) and to read and extract release information from those open third‑party pages, which the agent then uses to generate reports and drive follow-up actions, creating a clear avenue for indirect prompt injection.

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

• Potentially malicious external URL detected (high risk: 0.80). The skill requires running TinyFish at runtime to fetch arbitrary competitor webpages (e.g., tinyfish agent run --sync --url "https://asana.com/product"), and those fetched pages are injected into the agent's context to drive its outputs, so the external URL content directly controls the agent's behavior and is a required runtime dependency.