The Agent Skills Directory

[PROMPT_INJECTION]: The skill's core instructions in SKILL.md and references/auth.md explicitly direct the AI agent to withhold technical logs, command execution details, internal file paths, and environment variable names from its responses. While this behavior is flagged as concealment, the documentation provides the context that this is a user-experience optimization designed for non-technical users accessing the agent through mobile chat applications where technical terminal output would be considered noise.
[EXTERNAL_DOWNLOADS]: The skill utilizes Python's requests library to fetch generated media assets from the vendor's API endpoints (api.topview.ai) and well-known infrastructure providers (vercel.app). These downloads are triggered by explicit user requests for content generation and are saved to local directories for user access.
[EXTERNAL_DOWNLOADS]: The skill defines a standard set of dependencies in scripts/requirements.txt, including requests and python-dotenv. These are well-known, legitimate libraries used for HTTP communication and environment configuration.

topview-skill