The Agent Skills Directory

[CREDENTIALS_UNSAFE]: The skill manages a vendor-specific API key (PIXAZO_API_KEY). It instructs the agent to check for the key in environment variables or a specific local file (~/.pixazo/api-key). If not found, it prompts the user to provide it and subsequently saves it to the same local file using chmod 600 to restrict file permissions to the current user.
[EXTERNAL_DOWNLOADS]: The skill instructions direct the agent to fetch additional technical reference material from the vendor's official domain (https://www.pixazo.ai/models/fashn-vton.md) when detailed parameter schemas are required.
[COMMAND_EXECUTION]: The skill provides templates for the agent to generate and execute shell (curl), Python, and Node.js code to interact with the API endpoints hosted at gateway.pixazo.ai.
[INDIRECT_PROMPT_INJECTION]: The skill possesses an attack surface for indirect prompt injection as it processes external data.
Ingestion points: The agent fetches external documentation from the vendor's website and processes user-provided image URLs.
Boundary markers: None defined for the external documentation fetch.
Capability inventory: The skill has the ability to perform network requests and write to the local filesystem (~/.pixazo/api-key).
Sanitization: No specific sanitization or validation of the fetched markdown content is described.

fashn-vton