The Agent Skills Directory

[SAFE]: The skill instructions do not contain malicious patterns such as prompt injection, obfuscation, or unauthorized access to sensitive files. It performs standard documentation tasks using the AskUserQuestions tool to gather requirements.- [INDIRECT_PROMPT_INJECTION]: The skill exhibits an inherent surface for indirect prompt injection because it processes user-provided requirements to generate specifications.
Ingestion points: User input is gathered via the AskUserQuestions tool.
Boundary markers: The skill does not implement explicit delimiters for separating user-provided data from agent instructions.
Capability inventory: The skill is permitted to write documentation to the local specs/ directory.
Sanitization: No input sanitization is specified for the requirements provided by users.
Context: Given the limited capability to writing markdown files in a specific directory, this remains a standard risk for this type of tool and does not escalate the verdict.

feature-forge