paper-graph
Pass
Audited by Gen Agent Trust Hub on Jun 20, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill's operational flow involves executing its own internal Python scripts via the
uv runcommand to perform deterministic data processing and rendering tasks defined in a 13-step runbook. - [EXTERNAL_DOWNLOADS]: The skill fetches scientific paper metadata and full-text content from legitimate academic and technical domains, including Semantic Scholar (
api.semanticscholar.org), arXiv (arxiv.org), and PDF-to-Markdown conversion services (arxiv2md.org,markxiv.org). - [PROMPT_INJECTION]: The skill ingests untrusted content such as paper abstracts and conclusion excerpts from external sources. To mitigate potential indirect prompt injection attacks, the system uses structured delimiters and explicit instructions in its prompt templates to ensure the AI remains focused on the provided research context and follows a strict JSON or Markdown output format.
Audit Metadata