The Agent Skills Directory

[SAFE]: The skill implements a robust, structured workflow for requirements gathering and specification writing. It uses predefined templates and standard elicitation tools to ensure consistency and quality. The inclusion of 'MUST DO' constraints for security considerations and error handling demonstrates a security-conscious design for the generated documentation.- [INDIRECT_PROMPT_INJECTION]: The skill operates by ingesting user-provided requirements and writing them into markdown specification files. This process creates a theoretical attack surface for indirect prompt injection, where a user could provide malicious instructions that are subsequently saved into the project's documentation and later read by other AI agents.
Ingestion points: User input gathered through the AskUserQuestions tool and interview process defined in references/interview-questions.md.
Boundary markers: The output templates in references/specification-template.md do not utilize specific delimiters or 'ignore' instructions to isolate user-supplied content from the rest of the document.
Capability inventory: The skill utilizes file system writing capabilities to save documents to the specs/ directory.
Sanitization: No explicit sanitization or escaping of the user-provided text is performed before it is written to the specification files.

feature-forge