The Agent Skills Directory

[SAFE]: No malicious patterns, obfuscation, or unauthorized external communications were detected. The skill functions as a text-processing and file-management tool within the project scope.
[PROMPT_INJECTION]: The skill is designed to process untrusted user input (brain dumps) to generate prompts, which presents an indirect prompt injection surface.
Ingestion points: User-provided notes or existing prompt drafts (SKILL.md).
Boundary markers: None explicitly defined for the input notes to distinguish instructions from data.
Capability inventory: File system write operations to docs/prompts/ or docs/ (SKILL.md).
Sanitization: Employs an 'AI Self-Check' list requiring the agent to verify that no unauthorized instructions or 'slop' were added to the output.

prompt-generator