The Agent Skills Directory

Indirect Prompt Injection (LOW): The skill processes untrusted input from users via AskUserQuestion and writes it directly to the local filesystem.
Ingestion points: User responses provided through the AskUserQuestion tool during the interview loop.
Boundary markers: No explicit delimiters or instructions are provided to the agent to treat user answers as strictly passive data rather than instructions.
Capability inventory: The agent has the capability to create directories and write Markdown files (.md) to the local filesystem based on user input.
Sanitization: There is no evidence of sanitization or escaping of user input before it is interpolated into the generated Markdown files, which could allow a user to inject malicious formatting or instructions into the resulting knowledge base.

deep-interview