codex-slides-deck

Pass

Audited by Gen Agent Trust Hub on Jul 22, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructions direct the agent to execute a local script (../codex-slides/scripts/codex-slides.mjs) to handle deck processing and automation tasks.
  • [SAFE]: The workflows for project management, brand alignment, and design file manipulation use specialized tools that appear consistent with the vendor's ecosystem.
  • [SAFE]: The skill's surface for indirect prompt injection is documented based on its requirement to process user-provided content:
  • Ingestion points: Materials, design files, and images uploaded via upload_material, upload_design_file, and mark_edit_slide.
  • Boundary markers: The instructions do not define specific markers to delimit or ignore instructions within these files.
  • Capability inventory: Local CLI script execution and file modification tools.
  • Sanitization: No explicit sanitization or validation of the ingested content is specified in the skill files.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 22, 2026, 04:05 PM
Security Audit — agent-trust-hub — codex-slides-deck