The Agent Skills Directory

[SAFE]: The skill is composed of architectural documentation, workflow guides, and reference materials. It does not include any executable scripts or hidden logic, and all provided code snippets are illustrative patterns for API design and configuration.
[COMMAND_EXECUTION]: The documentation instructs the agent to utilize standard CLI tools for API lifecycle management, such as Spectral for linting, swagger-cli for validation, and openapi-generator-cli for code generation via npx. These operations are intended for developer-led workflows in local environments.
[EXTERNAL_DOWNLOADS]: The skill references numerous reputable Node.js packages (Spectral, Redocly, Prism, OpenAPI Generator) as part of its recommended tooling. These dependencies are well-known within the developer community and are used as intended for API validation, documentation, and testing.
[PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection as it ingests user-supplied API endpoints or resource definitions (SKILL.md) to generate specifications. The workflow lacks explicit boundary markers for untrusted data. However, the risk is mitigated by explicit security rules (SKILL.md) that mandate strict input validation, security schema definitions, and threat modeling, directing the agent toward safe data processing.

backend-openapi-documentation