lovstudio-fill-web-form

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

• Third-party content exposure detected (high risk: 1.00). The skill's mandatory Workflow Step 1 in SKILL.md requires using WebFetch(url) to retrieve and extract ALL fields from a user-provided web form URL (an arbitrary third‑party page) and then uses that extracted content to drive mapping, questions, and fill actions, which exposes the agent to untrusted external content that could carry indirect prompt injections.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill performs runtime WebFetch on user-provided form URLs (e.g., https://example.com/apply?formId=123) and injects the extracted form fields into the Agent sub-prompts, so fetched page content can directly control prompts used by the agent.