The Agent Skills Directory

[SAFE]: No malicious patterns or security risks were identified. The skill is instructional and guides the agent in generating architectural documentation.
[NO_CODE]: The skill consists entirely of Markdown documentation and instructions. It does not include any scripts (Python, JavaScript, shell, etc.) or executable files.
[DATA_EXFILTRATION]: There are no patterns suggesting unauthorized access to sensitive files or exfiltration of data to external domains. The skill operates solely on text inputs provided in the prompt context.
[PROMPT_INJECTION]: No attempts to bypass safety filters or override system instructions were identified. The instructions use standard language to define task workflow and quality bars.
[EXTERNAL_DOWNLOADS]: While external research and template tools are mentioned, they are referred to as platform-provided capabilities for gathering design patterns, not as external scripts or code to be downloaded.
[SAFE]: Indirect Prompt Injection Surface Analysis:
Ingestion points: Reads architectural requirements from external documents like user-flows and design-systems.
Boundary markers: Absent.
Capability inventory: The skill is restricted to text generation and does not have file-write or network-access capabilities.
Sanitization: None specified, but the absence of dangerous capabilities makes the attack surface non-exploitable.

authoring-wireframes