sprint-planning
Pass
Audited by Gen Agent Trust Hub on Mar 24, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill provides instructions to execute a local Python script (
generate_schematic.py) using theBashtool to create visual diagrams. This script is part of the project environment's skill structure. - [PROMPT_INJECTION]: The skill contains an indirect prompt injection surface through the diagram generation command. It instructs the agent to pass a user-supplied "diagram description" directly into a shell command. If the input contains shell metacharacters and the agent does not perform sanitization, it could lead to unintended command execution.
- Ingestion points: The text provided as a "diagram description" in
SKILL.md. - Boundary markers: The example uses double quotes around the description, which offers limited protection against shell-level injection.
- Capability inventory: The skill has access to the
Bashtool for local script execution. - Sanitization: No explicit sanitization or validation steps are defined in the instructions for the agent.
Audit Metadata