fetch-slides

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

• Third-party content exposure detected (high risk: 0.80). The skill fetches arbitrary public URLs (httpx.head/get and Playwright in fetch_slides.py and SKILL.md) and explicitly parses HTML from RevealJS presentations to extract slide text and links into slides_content.md, exposing the agent to untrusted, user-generated third‑party content that can influence downstream behavior.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill fetches arbitrary RevealJS presentation URLs at runtime (e.g., https://example.com/slides/) and loads them into Playwright with page.goto(...), thereby executing the remote page's JavaScript in a headless browser which the skill depends on to render PDFs.