The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses multiple shell commands to inspect the project environment, modify configurations, and trigger builds.
Evidence: Use of grep, find, sed, rm, xcodebuild, ./gradlew, and pod deintegrate across various phases (Phase 1, 5, and 7) to discover modules, build frameworks, and clean up CocoaPods artifacts.
[PRIVILEGE_ESCALATION]: The migration instructions guide the modification of security controls within the Xcode project environment.
Evidence: Phase 5.1 instructs the agent to disable Xcode's "User Script Sandboxing" (ENABLE_USER_SCRIPT_SANDBOXING = NO) using sed. This is a documented requirement for Gradle-based build phases to write to the project directory in Xcode 16+.
[DYNAMIC_EXECUTION]: The skill employs a dynamic helper script to process complex build metadata.
Evidence: Phase 5.1 uses python3 -c to execute an inline script that parses xcodebuild output in JSON format to identify and filter project schemes dynamically.
[EXTERNAL_DOWNLOADS]: The skill references and downloads dependencies from established organizations and well-known services.
Evidence: The migration mapping documentation references official GitHub repositories for Apple, Google, Firebase, and Kotlin (JetBrains) as sources for Swift Package Manager dependencies.
[INDIRECT_PROMPT_INJECTION]: The skill processes user-controlled project files which serve as inputs for its migration logic.
Ingestion points: Reads local configuration files including build.gradle.kts, Podfile, gradle.properties, and Kotlin source files.
Boundary markers: The skill does not define specific boundary markers for the ingested project data.
Capability inventory: The skill possesses the capability to modify project files and execute shell commands based on its analysis of the ingested data.
Sanitization: Employs rigid transformation logic (regex-based import replacement and targeted configuration modification) to mitigate the risk of arbitrary instruction execution.

kotlin-tooling-cocoapods-spm-migration