The Agent Skills Directory

[SAFE]: No malicious patterns detected. The skill provides standard coding guidance for Swift concurrency (async/await, Actors, TaskGroups) and testing frameworks (XCTest, Swift Testing).
[COMMAND_EXECUTION]: The skill includes instructions to run swift test and swift build in the verification phase. These are standard Swift development tools and are appropriate for the skill's primary purpose of supporting Swift engineering workflows.
[DATA_EXPOSURE]: Code examples use a standard placeholder domain (https://api.example.com) for networking demonstrations. No hardcoded credentials, sensitive file paths, or unauthorized network operations were found.
[INDIRECT_PROMPT_INJECTION]: The skill possesses a natural attack surface common to coding assistants (ingesting user code and suggesting build commands). However, it adheres to best practices by focusing on structured concurrency and testing, and no specific exploitable chains were identified.
Ingestion points: User-provided Swift code and development requests (SKILL.md).
Boundary markers: None explicitly defined for untrusted code, though instructions follow standard developer isolation patterns.
Capability inventory: Access to swift CLI for building and testing (SKILL.md).
Sanitization: Not applicable as the skill acts as a reference for developer-led workflows.

swift