The Agent Skills Directory

[CREDENTIALS_UNSAFE]: The skill implements a workflow for handling the Pixazo API key that includes checking environment variables and a dedicated local file (~/.pixazo/api-key). It correctly advises the agent to set restrictive file permissions (chmod 600) when saving the key, which protects the secret from other users on the system.- [EXTERNAL_DOWNLOADS]: The skill interacts with official vendor endpoints at gateway.pixazo.ai and provides example requests targeting audio assets hosted on Cloudflare R2 storage. These interactions are necessary for the skill's functionality and target reputable services.- [COMMAND_EXECUTION]: The instructions contain standard shell, Python, and Node.js examples for making HTTP requests to the Pixazo API. These examples are for implementation guidance and do not contain malicious payloads or system-altering commands.- [INDIRECT_PROMPT_INJECTION]: The skill ingests and processes JSON data from the Pixazo API gateway. The instructions guide the agent to extract specific result fields (like audio URLs) to display to the user, which is a standard pattern for API-driven skills.

gemini-voice