The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes shell commands such as grep, npm run, and npx tsc to identify and run the appropriate type-checking scripts within a project.
[INDIRECT_PROMPT_INJECTION]: The skill possesses an attack surface by reading and acting upon project-controlled files and compiler output.
Ingestion points: Reads command names from package.json and parses error descriptions from tsc output.
Boundary markers: None identified.
Capability inventory: Shell command execution (npm, npx, grep) and file system access to apply fixes.
Sanitization: No explicit sanitization or validation is mentioned for project-defined script names or error log contents.
[DYNAMIC_EXECUTION]: Shell commands are dynamically constructed by prefixing execution logic with environment setup commands (nvm use <version>) where the version is derived from project configuration files.

typescript-check