ppt-creator
Warn
Audited by Gen Agent Trust Hub on Feb 23, 2026
Risk Level: MEDIUMCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- Dynamic Execution (MEDIUM): The orchestration workflow described in 'ORCHESTRATION_DATA_CHARTS.md' and 'ORCHESTRATION_OVERVIEW.md' instructs the AI to generate a Python script ('generate_charts.py') and execute it using the system interpreter or 'uv'. This runtime code generation and execution is a medium-risk activity.
- Indirect Prompt Injection (LOW): The skill processes user-supplied data and instructions to build presentations, creating a vulnerability surface.
- Ingestion points: Data ingestion occurs during Stage 8b via user-uploaded CSV/Excel files and via the intake questionnaire in 'INTAKE.md'.
- Boundary markers: No explicit boundary markers or 'ignore' instructions for embedded data were found in the analyzed files.
- Capability inventory: The skill executes shell commands (via 'python' or 'uv') and utilizes a 'document-skills:pptx' subagent.
- Sanitization: The provided 'scripts/chartkit.py' performs basic structural checks on data but does not sanitize content against malicious injection.
Audit Metadata