The Agent Skills Directory

[PROMPT_INJECTION]: No evidence of prompt injection or instructions to bypass safety guardrails. The skill defines strict design constraints for the agent to follow, such as frequency gates for animation, which are benign instructional boundaries.\n- [DATA_EXFILTRATION]: No network operations, sensitive file path access, or credential harvesting patterns were detected. The skill operates as a read-only advisor for user-provided UI code.\n- [REMOTE_CODE_EXECUTION]: No external package installations or remote script downloads are referenced. All logic is derived from local markdown references.\n- [INDIRECT_PROMPT_INJECTION]: The skill processes user-provided UI source code and design specifications (Ingestion points). It lacks dangerous capabilities such as file writing, network access, or command execution (Capability inventory), rendering the ingestion surface safe despite a lack of explicit boundary markers.\n- [DYNAMIC_EXECUTION]: The code snippets provided in the reference files are for educational and illustrative purposes only and are not executed by the skill environment.\n- [SAFE]: The skill adheres to the 'review-only' protocol established in its own documentation, limiting its scope to analysis and planning rather than autonomous mutation of the system.

motion-craft