The Agent Skills Directory

[DATA_EXPOSURE]: The skill instructions direct the agent to retrieve and store the PIXAZO_API_KEY. It looks for the key in environment variables or the local file ~/.pixazo/api-key. If found, it saves the key with restrictive permissions (chmod 600) in the user's home directory. This is standard practice for local credential management.
[EXTERNAL_DOWNLOADS]: The skill fetches model schemas and detailed parameter information from the vendor's official documentation at https://www.pixazo.ai/models/runway.md to ensure correct API interactions.
[COMMAND_EXECUTION]: The documentation references an npx command for installing additional skills from the Pixazo-AI/skills repository. This is a standard installation method for vendor-provided extensions.
[INDIRECT_PROMPT_INJECTION]: The skill ingests external content from the vendor's model documentation site. While this content influences how the agent structures API calls, the data is retrieved from the vendor's own infrastructure.
Ingestion points: External model reference documentation fetched from https://www.pixazo.ai/models/runway.md (SKILL.md).
Boundary markers: None specified for the external documentation content.
Capability inventory: Performs network requests via curl, Python requests, and Node.js fetch to gateway.pixazo.ai (SKILL.md).
Sanitization: Not explicitly defined for the fetched documentation content.

runway