timeline-generator-scott-margetts
Pass
Audited by Gen Agent Trust Hub on Jun 26, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: The skill ingests data from external sources such as CSV files and monitored Outlook emails, which creates an indirect prompt injection surface. Instructions hidden in these inputs could potentially influence the agent's actions.
- Ingestion points: Processes matter-plan-builder CSV exports and monitors Outlook emails (SKILL.md).
- Boundary markers: No specific delimiters or "ignore instructions" warnings are provided to separate user data from instructions.
- Capability inventory: The skill can update SharePoint lists, Planner milestones, and draft Outlook emails (SKILL.md).
- Sanitization: No validation or sanitization of inputs is mentioned before the data is used for calculations or visualization.
- [COMMAND_EXECUTION]: The skill dynamically generates HTML and JavaScript for an interactive canvas-based Gantt chart. Generating executable scripts based on external data (like task names from emails) presents a surface for dynamic code injection or XSS within the agent's visualization context.
Audit Metadata