The Agent Skills Directory

[PROMPT_INJECTION]: No evidence of attempts to override agent behavior, bypass safety filters, or extract system prompts was found. The use of 'CRITICAL' in the metadata is used to define usage priority for the agent, not to override core instructions.
[DATA_EXFILTRATION]: No sensitive file paths, credential harvesting, or unauthorized network operations were detected. The skill mentions standard application data directories for persistence purposes.
[REMOTE_CODE_EXECUTION]: No patterns involving remote script execution, such as 'curl | bash', or dynamic code execution were found.
[COMMAND_EXECUTION]: The skill provides Rust code snippets for application logic and does not contain any suspicious system command execution or privilege escalation attempts.
[OBFUSCATION]: The content is clear and uses standard Markdown and Rust syntax. No hidden characters, Base64 encoding of commands, or homoglyph attacks were detected.
[PERSISTENCE]: While the skill mentions saving and loading application state (window geometry, user state), these are standard functional requirements for desktop applications and do not constitute malicious persistence mechanisms.

robius-app-architecture