The Agent Skills Directory

[COMMAND_EXECUTION]: The skill leverages the Bash(curl *) tool to make POST requests to the each::labs API endpoint (https://eachsense-agent.core.eachlabs.run). This is the intended primary purpose of the skill.
[EXTERNAL_DOWNLOADS]: The skill references external resources from eachlabs.ai and eachlabs.run for API interactions and documentation. These are the official domains for the service provider associated with the skill.
[INDIRECT_PROMPT_INJECTION]: The skill possesses an attack surface for indirect prompt injection as it ingests untrusted user input (text descriptions of sounds) and external image URLs to be processed by an AI model.
Ingestion points: User messages in the content field and image_url values provided in API payloads.
Boundary markers: No explicit delimiters or instructions to ignore embedded commands are present in the provided templates.
Capability inventory: The skill uses Bash(curl *) for network requests and WebFetch for data ingestion.
Sanitization: No specific sanitization or filtering logic is defined within the skill instructions.

sound-effects