data-export-pdf

Pass

Audited by Gen Agent Trust Hub on Jul 16, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The Troubleshooting section provides a code snippet that uses the pip package manager to download and install the 'reportlab' library from the official Python Package Index (PyPI). This is a standard procedure for environment setup.
  • [COMMAND_EXECUTION]: The skill includes a troubleshooting example that uses the subprocess module to programmatically execute pip install commands. This allows the agent to manage its own dependencies if the required library is missing.
  • [DATA_EXFILTRATION]: The skill processes biological data objects (Anndata) to extract metadata and statistics for report generation. This represents an indirect prompt injection surface where untrusted data is ingested into the agent context.
  • Ingestion points: The create_analysis_report function in SKILL.md ingests data from the adata object's observations (adata.obs) and metadata.
  • Boundary markers: None present in the code snippets provided.
  • Capability inventory: The skill performs local file system writes to save the generated PDF and temporary plot images.
  • Sanitization: The skill does not explicitly sanitize data before passing it to reportlab.platypus.Paragraph objects, which may interpret internal markup tags.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 16, 2026, 07:45 PM
Security Audit — agent-trust-hub — data-export-pdf