The Agent Skills Directory

[SAFE]: The skill provides documentation and illustrative code snippets for front-end performance optimizations, such as image compression, bundle reduction, and GPU acceleration. All code provided is for demonstration and does not contain malicious logic.- [PROMPT_INJECTION]: No behavioral overrides, safety bypasses, or adversarial instructions were found. The use of instructional terms like 'CRITICAL' and 'IMPORTANT' is contextually appropriate for performance engineering.- [DATA_EXFILTRATION]: The skill does not perform network requests to external domains, nor does it access sensitive local file paths or credentials.- [REMOTE_CODE_EXECUTION]: There are no patterns involving the download or execution of remote scripts, unverifiable dependencies, or dynamic code generation.

optimize