The Agent Skills Directory

[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection attacks because it processes untrusted user input to generate Mermaid diagram code.\n
Ingestion points: User-provided descriptions used in the diagram generation process (SKILL.md).\n
Boundary markers: Absent; there are no instructions for the agent to use delimiters or to ignore instructions embedded within the user data.\n
Capability inventory: The skill uses the Bash tool to perform file-writing operations.\n
Sanitization: Absent; the instructions do not require the agent to escape shell metacharacters or validate the input before processing it.\n- [COMMAND_EXECUTION]: There is a risk of command injection through the file-saving logic. The skill instructs the agent to save diagrams using echo via the Bash tool. If the generated diagram content, which is derived from user input, contains shell-sensitive characters like backticks, semicolons, or command substitutions, it could lead to the execution of unintended commands.

diagram-gen