felo-content-to-slides

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

• Third-party content exposure detected (high risk: 0.90). The SKILL.md and README explicitly instruct the agent to fetch arbitrary webpage URLs and YouTube videos (via --url/--video using felo-web-fetch and felo-youtube-subtitling) and then read/interpret that untrusted third-party content to generate PPTs, so external page/video content could inject instructions that influence generation and downstream actions.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill fetches arbitrary webpage or YouTube content at runtime (e.g., --url "https://example.com/article" or --video "https://www.youtube.com/watch?v=ID"), and that fetched content is injected into the PPT-generation prompts so the external URLs directly control the model's instructions.