The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes a TypeScript wrapper (scripts/module_architect_wrapper.ts) to execute PowerShell scripts using child_process.spawn. This mechanism is used to run tasks such as module scaffolding (scaffold_module.ps1), manifest creation (create_manifest.ps1), and function validation (validate_functions.ps1). While the wrapper uses arguments rather than shell strings, it represents a core capability of executing shell-level commands.
[INDIRECT_PROMPT_INJECTION]: The skill includes functionality to ingest and analyze user-provided PowerShell scripts to extract function signatures, perform syntax checks, and refactor code into modules. Maliciously crafted scripts could attempt to embed instructions designed to influence the agent's logic during these analysis phases.
Ingestion points: The skill reads file content in scripts/create_manifest.ps1 (via Get-Content) and scripts/validate_functions.ps1 (via [System.Management.Automation.Language.Parser]::ParseFile) to perform automated analysis.
Boundary markers: The instructions do not define specific boundary markers or "ignore embedded instructions" warnings when processing these external files.
Capability inventory: The skill possesses file system access (read, write, delete) and the ability to execute PowerShell commands through the pwsh environment.
Sanitization: While basic validation is performed on module names (e.g., in scripts/scaffold_module.ps1), the content of analyzed scripts is processed via regular expressions and AST parsing without explicit sanitization against injection attacks.

powershell-module-architect