sciatlas-trend-report

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

• Third-party content exposure detected (high risk: 0.85). The required runtime workflow ingests outsider-authored free text from SciAtlas search-papers results (public paper metadata/abstracts/refs returned by the external SciAtlas service) into the agent’s LLM context for trend analysis.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill's bootstrap step instructs running pip install "git+https://github.com/zjunlp/SciAtlas.git#subdirectory=sciatlas" at runtime, which fetches and installs remote code (executing it) and is a required dependency for the sciatlas CLI used by the skill.