The Agent Skills Directory

[COMMAND_EXECUTION]: The skill includes local Python scripts (scripts/generate_api.py and scripts/validate_api.py) designed to be executed via the Bash tool to scaffold and validate API design documents.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to process and review external API designs, which presents a surface for indirect prompt injection if the content being reviewed contains malicious instructions.
Ingestion points: User-provided API designs or existing specification files read during the 'Review' or 'Improve' tasks defined in SKILL.md.
Boundary markers: The skill instructions do not specify the use of delimiters or 'ignore' instructions to isolate untrusted API data from the agent's core instructions.
Capability inventory: The skill has access to Write, Edit, and Bash tools, which could be targeted by an injection attack.
Sanitization: No specific content validation or sanitization is performed on the data being reviewed.
[EXTERNAL_DOWNLOADS]: The skill references documentation and best practices from official and trusted sources, including Microsoft's GitHub repository and the GraphQL Foundation's learning resources.

api-designer