papers-skill

Pass

Audited by Gen Agent Trust Hub on Jun 16, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill installs the httpx, arxiv, and PyMuPDF packages from the official Python package registry (PyPI) to facilitate network requests and PDF processing.
  • [COMMAND_EXECUTION]: The agent runs a bundled command-line interface script (scripts/papers.py) to orchestrate its search, metadata retrieval, and file handling capabilities.
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it ingest and processes text from external academic papers without sanitization.
  • Ingestion points: The read subcommand in scripts/papers.py extracts text directly from local PDF files into the agent's context.
  • Boundary markers: There are no explicit delimiters or instruction-override warnings used to wrap the text extracted from PDFs.
  • Capability inventory: The skill possesses the ability to write to the file system (download), execute subprocesses (the bundled CLI), and perform network operations (API calls).
  • Sanitization: The extracted text is returned as a raw string without filtering for potential adversarial instructions hidden in the document content.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 16, 2026, 08:49 AM
Security Audit — agent-trust-hub — papers-skill