The Agent Skills Directory

[SAFE]: The skill consists of natural language instructions for a document-drafting workflow. It does not contain executable code, scripts, or system commands that could be used for malicious purposes.
[SAFE]: No network operations, credential harvesting, or unauthorized file access patterns were detected. The skill uses platform-provided document tools for its intended purpose of creating Word documents.
[INDIRECT_PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection as it is designed to process user-provided NDA documents for review.
Ingestion points: User-supplied cooperation backgrounds and third-party NDA text for review (SKILL.md).
Boundary markers: The instructions do not define specific delimiters to separate user data from system instructions, though the agent is instructed to treat the input as a reference for review.
Capability inventory: The skill uses internal document generation tools to create Word files.
Sanitization: No specific sanitization or filtering logic is mentioned for the processed text.
[SAFE]: The presence of an indirect prompt injection surface is a standard characteristic of document-processing agents and, in this context, is not assessed as a threat given the skill's specific task and lack of sensitive capabilities.

nda-drafter