The Agent Skills Directory

[SAFE]: The skill operates entirely within the project workspace, referencing local files for design standards and using standard development tools like dart analyze and ng lint. No external network operations, remote code downloads, or sensitive file access patterns were detected.
[PROMPT_INJECTION]: The skill uses strong language such as "Iron Law" and "Hard Policy" to guide the agent's code generation and review process. While prescriptive, these instructions are aimed at maintaining design consistency and do not attempt to bypass safety filters or override system-level instructions.
[DATA_EXFILTRATION]: There is no evidence of data exfiltration. The skill uses tools like Read, Glob, Grep, Write, and Edit to perform UI audits and implement design tokens within source code files. It does not access credentials, environment variables, or other sensitive user data.
[REMOTE_CODE_EXECUTION]: The skill does not execute any remote scripts or download external dependencies at runtime. All linting and analysis commands mentioned are standard framework-specific tools.

design-system