adhx

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

• Third-party content exposure detected (medium risk: 0.30). The workflow ingests outsider-authored free text from user-provided X/Twitter posts by parsing a shared URL and fetching post content via the ADHX API endpoint (https://adhx.com/api/share/tweet/{username}/{statusId}), which can include arbitrary text payloads.

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

• Potentially malicious external URL detected (high risk: 0.80). The skill issues runtime curl requests to https://adhx.com/api/share/tweet/{username}/{statusId} (shown in multiple examples such as https://adhx.com/api/share/tweet/dgt10011/2020167690560647464 and https://adhx.com/api/share/tweet/handle/123), fetching external JSON that is injected into the agent context and directly controls the skill's output.