The Agent Skills Directory

[PROMPT_INJECTION]: No evidence of malicious instructions, jailbreak attempts, or safety filter bypasses was found. The instructions are purely technical and instructional.\n- [DATA_EXFILTRATION]: No unauthorized network operations or access to sensitive local files (e.g., credentials, SSH keys) were detected. The skill uses standard Android APIs for camera data and inter-process communication via the Wear OS Data Layer API.\n- [EXTERNAL_DOWNLOADS]: The skill references official dependencies and documentation from trusted domains such as developer.android.com. These represent standard, legitimate technical resources.\n- [SAFE]: Indirect Prompt Injection Surface: The skill includes guidance for ingesting and processing camera frame data via ML Kit. While this represents a data ingestion point, the analysis focuses on coordinate transformation and UI rendering best practices, and the skill does not expose exploitable capabilities such as arbitrary file writes or network transmission based on the processed data.\n- [SAFE]: Metadata Consistency: The skill metadata attributes the content to Google LLC, which is consistent with the subject matter. No deceptive or malicious metadata poisoning was identified.

camerax