The Agent Skills Directory

[SAFE]: The skill consists of technical documentation, best practices, and code examples for SwiftUI development. No malicious patterns, obfuscation, or unauthorized access attempts were detected across the provided files.
[DATA_EXPOSURE_AND_EXFILTRATION]: The skill contains multiple references to sosumi.ai (e.g., in references/platform-and-sharing.md and references/preview-isolation.md). These are used as informational links to proxy Apple developer documentation and WWDC sessions. They do not involve sensitive data transmission or automated network operations.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze user-provided SwiftUI source code to provide refactoring advice and architectural guidance.
Ingestion points: User-supplied SwiftUI views, models, and architecture queries (described in SKILL.md workflow).
Boundary markers: None identified.
Capability inventory: The skill provides instructions for the agent to generate advice and code refactors. It does not include tools or scripts for shell execution, network activity, or persistent file modifications.
Sanitization: None identified. Given the lack of actionable capabilities that could be abused by an attacker, the ingestion surface is considered safe.
[COMMAND_EXECUTION]: The provided files contain Swift code examples, but no shell commands or instructions for the agent to execute arbitrary system commands were found.

swiftui-patterns