The Agent Skills Directory

[SAFE]: The skill content was analyzed for malicious patterns, including prompt injection, data exfiltration, and obfuscation. No such patterns were identified. The skill is composed of markdown instructions and JSON metadata.
[EXTERNAL_DOWNLOADS]: The skill includes links to official Apple developer documentation (developer.apple.com) and WWDC videos. These are recognized as trusted, well-known services, and the references are used for legitimate technical guidance.
[INDIRECT_PROMPT_INJECTION]: The skill contains instructions for processing user-provided source code to perform modernization tasks. While this establishes an ingestion surface for untrusted data, the skill includes explicit safety guidelines, such as requiring user confirmation for risky changes and providing actionable TODOs instead of silent failures, which mitigates the risk of unintended behavior.
[COMMAND_EXECUTION]: The UPSTREAM.json file contains a metadata field export_command which describes the command used by the author to generate the skill assets from Apple's Xcode environment. This is documented for provenance and is not executed by the agent on the user's system.

app-resizability