The Agent Skills Directory

[PROMPT_INJECTION]: The skill processes untrusted user input from the conversation context to generate HTML artifacts and diagrams. This creates a surface for indirect prompt injection where an attacker could influence the conversation to include malicious scripts in the generated visual aids.
Ingestion points: The agent's conversation history ('current topic of conversation') as described in SKILL.md.
Boundary markers: None provided to distinguish between user content and visualization templates.
Capability inventory: Instructions to generate files and execute shell commands (Bash) are present in SKILL.md.
Sanitization: The skill lacks instructions for escaping or sanitizing conversation data before it is rendered into HTML.
[COMMAND_EXECUTION]: The skill uses the Bash tool to execute the open command (e.g., Bash(open path/to/show-me-{description}.html)). While this is intended to display visual artifacts to the user, it involves shell command execution tied to dynamically generated filenames.

show-me