The Agent Skills Directory

[COMMAND_EXECUTION]: The utility script scripts/preview-script.sh is designed to execute local shell or Python scripts to capture and preview their output. This is a fundamental capability required for testing Raycast commands.
[DYNAMIC_EXECUTION]: The metadata validation utility scripts/check-raycast-script-metadata.sh executes an embedded Python script via stdin (python3 -) to perform regex-based analysis of script headers.
[INDIRECT_PROMPT_INJECTION]: The skill interacts with external script files in the development workspace, which represents a potential attack surface for indirect prompt injection if those files contain malicious instructions meant to influence the agent during analysis or preview.
Ingestion points: File reading occurs in scripts/check-raycast-script-metadata.sh, and execution occurs in scripts/preview-script.sh.
Boundary markers: No specific delimiters or safety warnings are implemented for processing script content.
Capability inventory: The skill provides capabilities for local file read access and local code execution.
Sanitization: No sanitization or sandboxing is applied to the script content prior to analysis or execution.
[EXTERNAL_DOWNLOADS]: The README.md file contains installation instructions referencing the author's official repository via the skills CLI and npx. These are legitimate vendor-owned installation paths.

build-raycast-script-command