theharvester

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

• Third-party content exposure detected (high risk: 0.95). TheHarvester’s runtime workflow fetches and parses public OSINT results from outsider-authored web sources (e.g., Google/Bing/DuckDuckGo/LinkedIn/Shodan/VirusTotal) and then ingests the scraped/free-text output into the agent via the generated JSON/stdout that the Python wrapper reads and parses.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill's installation/run steps explicitly fetch and execute remote code (e.g., "git clone https://github.com/laramies/theHarvester.git" and "docker pull ghcr.io/laramies/theharvester" followed by running python3 theHarvester.py / docker run), so these runtime URLs pull external code that is executed.