The Agent Skills Directory

[COMMAND_EXECUTION]: The skill's workflow includes instructions to use npx @redocly/cli for linting OpenAPI specifications and npx @stoplight/prism-cli for running mock servers. These are industry-standard tools for API development and are consistent with the skill's primary purpose.
[EXTERNAL_DOWNLOADS]: The use of npx implicitly involves fetching packages from the official npm registry. As this targets a well-known service for standard development tasks, it is considered a safe and expected operation.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process domain requirements and business logic provided by users to generate API specifications.
Ingestion points: Business requirements and data models provided during the 'Analyze domain' phase.
Boundary markers: The instructions do not explicitly define delimiters for user-provided content.
Capability inventory: The agent has the capability to write YAML files (openapi.yaml) and execute shell commands (npx).
Sanitization: No explicit sanitization or instruction-ignoring prompts are present for the processed domain data. However, the risk is inherent to the specialized architectural task and does not indicate malicious intent.

api-designer