The Agent Skills Directory

[SAFE]: The skill is designed to facilitate the creation of architectural documentation through a structured, interactive process. No malicious code, command execution, credential harvesting, or unauthorized network operations were detected.
[PROMPT_INJECTION]: The skill uses behavioral constraints like 'DISABLE ALL EFFICIENCY OPTIMIZATIONS' and 'MANDATORY STEP-BY-STEP EXECUTION' to ensure the agent follows the prescribed interactive workflow. These are functional instructions to maintain quality and do not constitute a bypass of safety guidelines or an attempt to override system prompts.
[INDIRECT_PROMPT_INJECTION]: The skill identifies vulnerability surfaces where it processes untrusted data from project documents (such as prd.md) and user-provided repository analysis to inform architectural decisions. This is inherent to the skill's primary purpose.
Ingestion points: Reads docs/prd.md and analyzes user-provided starter templates or repository files (referenced in SKILL.md and the templates folder).
Boundary markers: Absent. The skill does not explicitly use delimiters to separate untrusted document content from its internal instructions.
Capability inventory: The skill performs file system operations to read templates and write documentation to the docs/ directory. It does not possess capabilities for arbitrary code execution or network exfiltration.
Sanitization: Absent. Content from external documents is summarized and incorporated based on the agent's reasoning without explicit sanitization routines.

create-architecture-doc