The Agent Skills Directory

[SAFE]: The skill's behavior is consistent with its documentation. It functions as a UI component library and generative design tool without any hidden malicious logic.
[EXTERNAL_DOWNLOADS]: The project references and downloads resources from trusted domains and well-known services:
Fetches the core library from the vendor's domain daub.dev.
Utilizes Lucide icons and other utilities from standard CDNs including unpkg.com and jsdelivr.net.
Uses Google Fonts for enhanced typography.
[COMMAND_EXECUTION]: The allowed-tools section includes Bash, which is used for standard development operations like package installation (npm install). No high-risk or arbitrary command execution patterns were found in the skill's instructions.
[REMOTE_CODE_EXECUTION]: The skill uses a remote MCP server at daub.dev to assist in generating UI specifications. This is a legitimate extension of the library's functionality and does not involve executing untrusted scripts on the host system.
[DATA_EXFILTRATION]: No patterns indicating the exfiltration of sensitive information (such as credentials, environment variables, or SSH keys) were detected. The Figma integration uses a secure OAuth flow handled by the vendor's API proxy.
[PROMPT_INJECTION]: The skill's instructions focus on providing technical reference data and usage patterns. There are no attempts to bypass safety filters or manipulate the agent's core behavioral constraints.

daub-ui