requesthunt

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

• Third-party content exposure detected (medium risk: 0.30). The requesthunt skill aggregates user feedback and feature requests from public social platforms and issue trackers such as Reddit, X, and GitHub, allowing outsiders to submit untrusted text that the agent reads and analyzes.

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

• Potentially malicious external URL detected (high risk: 0.70). The skill instructs the agent to install a CLI tool via curl https://requesthunt.com/cli | sh and fetch binaries from an external GitHub repository during setup, which constitutes a weak/unknown/suspicious runtime external dependency.