analyzing-malware

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

• Third-party content exposure detected (medium risk: 0.30). The skill documentation’s “Reading External Sources” section shows that runtime execution would use curl to fetch arbitrary external pages (Markdown/extracted full text) and ingest their free text via an outsider-authored URL, which is an indirect prompt-injection exposure path.

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

• Potentially malicious external URL detected (high risk: 0.80). The skill instructs at runtime to fetch arbitrary external pages via curl to "https://defuddle.md/", which would retrieve remote content that can be injected into the agent's context and thus directly control prompts or analysis output.