skills/wyattowalsh/agents/trafilatura/Gen Agent Trust Hub

trafilatura

Pass

Audited by Gen Agent Trust Hub on Jul 7, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill wraps the trafilatura CLI using Python scripts. Analysis of extract_url.py, list_urls.py, and doctor.py shows that all system commands are executed using subprocess.run with list-based arguments and without shell invocation, effectively mitigating shell injection risks.\n- [DATA_EXFILTRATION]: No unauthorized data exfiltration patterns were detected. The skill's primary function is to fetch and clean web content, with outputs directed to stdout or a designated user download directory.\n- [PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection as it processes untrusted data from external URLs.\n
  • Ingestion points: Web content is ingested via extract_url.py and list_urls.py.\n
  • Boundary markers: The skill utilizes a JSON envelope for structured output in several modes to help separate metadata from content.\n
  • Capability inventory: The skill has access to the Bash and Write tools, and executes subprocesses via the provided Python scripts.\n
  • Sanitization: No explicit sanitization or filtering of the extracted article text is performed before presentation to the agent.\n
  • This risk is typical for scraping tools and is addressed through the skill's instructional framework and 'stdout-first' protocol.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 7, 2026, 06:43 AM
Security Audit — agent-trust-hub — trafilatura