The Agent Skills Directory

[SAFE]: The skill consists of technical documentation and code examples for implementing accessibility in Flutter applications. The content is educational and instructional in nature.
[PROMPT_INJECTION]: No attempts to override agent behavior, bypass safety guidelines, or extract system prompts were detected. The instructional language is focused on coding best practices.
[DATA_EXFILTRATION]: No sensitive data exposure or exfiltration patterns were found. The skill does not access credentials, private keys, or environment files. Network activity is limited to referencing official documentation from well-known sources like Flutter and W3C.
[REMOTE_CODE_EXECUTION]: The skill does not perform remote script execution or download unverifiable packages. It mentions standard development tools and lints like flutter_lints and very_good_analysis in a descriptive context.
[OBFUSCATION]: No obfuscated content, such as Base64-encoded commands, zero-width characters, or homoglyphs, was identified.
[COMMAND_EXECUTION]: The skill suggests a grep command for CI pipelines to detect anti-patterns, which is a standard quality assurance practice and not a malicious payload.

accessibility-as-code