The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill fetches model schemas, parameter names, and allowed values from an external URL: https://www.pixazo.ai/models/longcat-image.md.
[INDIRECT_PROMPT_INJECTION]: The skill processes external markdown content from a remote URL to determine its execution parameters, creating a surface for indirect prompt injection.
Ingestion points: Remote documentation fetched from https://www.pixazo.ai/models/longcat-image.md as specified in the 'Full reference' section of SKILL.md.
Boundary markers: None; the agent is instructed to load and follow the external content for exact parameter names and accepted values.
Capability inventory: The skill performs file writes (saving API keys to ~/.pixazo/api-key), modifies file permissions (chmod), and makes network requests (calling the Pixazo API gateway).
Sanitization: There is no instruction to sanitize or validate the content fetched from the external markdown file before use.
[COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands to manage file permissions (chmod 600) and provides an npx command for installing related skills from the Pixazo-AI repository.

longcat-image