The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses shell commands to diagnose and resolve issues, including swift build, swift package, and rm. Specifically, it includes instructions to delete the SwiftPM cache directory located at ~/Library/Caches/org.swift.swiftpm, which involves file operations outside the immediate project scope.
[COMMAND_EXECUTION]: The skill recommends the use of the --disable-sandbox flag with swift build as a resolution for macro target failures. This flag bypasses the default Swift package sandbox, potentially allowing third-party build plugins or macros to perform unauthorized network or filesystem operations.
[INDIRECT_PROMPT_INJECTION]: The skill has a defined attack surface for indirect prompt injection via project configuration files.
Ingestion points: The agent is instructed to read and analyze Package.swift and Package.resolved files from the user's project (SKILL.md).
Boundary markers: The instructions do not define clear delimiters or specific warnings to ignore natural language instructions that might be embedded within the project files.
Capability inventory: The agent has the capability to execute shell commands such as swift build, swift package, and file system deletions (rm).
Sanitization: There is no evidence of sanitization or validation of the manifest file contents before they are processed by the agent's logic.

axiom-resolve-spm