The Agent Skills Directory

[PRIVILEGE_ESCALATION]: The skill instructs the agent to use sudo to install system dependencies (sudo apt install graphviz). Execution of commands with superuser privileges allows for potential system-wide modifications and unauthorized access if exploited.
[COMMAND_EXECUTION]: The skill uses the Bash tool to perform software installations (choco, apt, brew, pip) and to execute Python scripts. This provides a direct interface for executing arbitrary commands on the host system.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user data to generate diagrams, creating an attack surface for indirect prompt injection.
Ingestion points: User-provided lists for steps and blocks, as well as raw Graphviz dot_code strings.
Boundary markers: The instructions do not define delimiters or warnings to ignore instructions embedded within the diagram data.
Capability inventory: The skill has access to Bash, Write, and Read tools, and modifies sys.path to execute local Python modules.
Sanitization: There is no evidence of sanitization or validation for the dot_code or component labels before they are processed by the rendering engine.

patent-diagram-generator