tailor-to-venue

Pass

Audited by Gen Agent Trust Hub on Jul 7, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes local Python scripts (venue_diff.py, page_budget.py, anon_sweep.py) to analyze LaTeX files. These scripts are self-contained and use only the Python standard library.
  • [EXTERNAL_DOWNLOADS]: The skill instructs the agent to fetch live data from Call for Papers (CFP) URLs to verify requirements. This is a legitimate functional requirement for ensuring the accuracy of page limits and deadlines.
  • [PROMPT_INJECTION]: The skill processes untrusted input from LaTeX source files and external CFP websites. While this represents a surface for indirect prompt injection, it is handled as a standard data processing task without identified malicious patterns.
  • [SAFE]: No obfuscation, hardcoded credentials, or persistence mechanisms were detected. The scripts include safety measures such as recursion depth limits in the LaTeX parser and a restricted custom YAML loader for venue profiles.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 7, 2026, 01:03 PM
Security Audit — agent-trust-hub — tailor-to-venue