The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses curl to perform various GET requests to the Omni REST API for model discovery and metadata inspection. Since these operations are directed at the vendor's platform (omniapp.co) and align with the skill's documented purpose, they are considered standard functionality and safe.
[EXTERNAL_DOWNLOADS]: The skill retrieves an openapi.json file from the vendor's domain to discover API capabilities. This is a common and safe pattern for API-integrated tools.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it retrieves and processes structured data (JSON/YAML) and metadata (such as ai_context instructions) from external API endpoints.
Ingestion points: Data enters the agent's context through several API endpoints described in SKILL.md, including /api/v1/models and topic inspection routes.
Boundary markers: The instructions do not define specific delimiters or instructions to ignore commands that might be embedded in the API responses.
Capability inventory: The skill has the capability to execute network requests via curl based on model information.
Sanitization: There is no mention of sanitization or validation of the retrieved API data before it is presented to or processed by the AI agent.

omni-model-explorer