slide-creator

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

• Third-party content exposure detected (high risk: 0.90). The skill explicitly instructs the agent to fetch and parse user-provided web URLs (via "web_fetch" / "When the user provides web URLs, use web_fetch to extract design cues" in SKILL.md and the Reference Extraction Protocol in references/art-direction.md), so it ingests untrusted public web content and uses that content to drive style selection and subsequent build/export actions.

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

• Potentially malicious external URL detected (high risk: 0.80). The skill explicitly calls for using web_fetch on user-provided web URLs (i.e., any https://... URL supplied as a style reference) at runtime to extract page copy/colors/layout and use those cues to drive the agent's art-direction prompts and generated CSS, so remote page content directly controls the agent's instructions.

MEDIUM W021: Hidden or invisible Unicode characters detected (potential obfuscation or prompt injection).

• Hidden Unicode characters detected (1 type(s) found)