paper-poster-html

Pass

Audited by Gen Agent Trust Hub on Oct 24, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external sources, creating an attack surface for indirect prompt injection.
  • Ingestion points: Phase 0 fetches venue specifications from the web (WebSearch/WebFetch) and Phase 1 reads user-provided paper source files (.tex/PDF).
  • Boundary markers: Absent. There are no explicit delimiters or instructions to ignore embedded commands in the ingested content.
  • Capability inventory: The skill can execute shell commands via Bash and orchestrate Python scripts using subprocess.run.
  • Sanitization: The skill uses an ascii_safe helper for terminal output but lacks content sanitization for fetched specifications or paper text before processing.
  • [COMMAND_EXECUTION]: The skill uses subprocess calls to execute internal logic and external utilities as part of its primary functionality.
  • Evidence: scripts/run_gates.py uses subprocess.run to execute the individual checking scripts; scripts/_posterly/verify_final.py uses subprocess.check_output to run pdfinfo for PDF validation.
  • [EXTERNAL_DOWNLOADS]: The skill references and downloads resources from well-known external services.
  • Evidence: Downloads tex-svg.js from cdn.jsdelivr.net for MathJax rendering and installs Playwright/Chromium dependencies at runtime.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 24, 2026, 09:43 AM