The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest human input through the ask and ask_secret functions in the generated scripts. This input is then used to populate environment variables and GitHub secrets.
Ingestion points: template.sh uses the read command to capture interactive user input at runtime.
Boundary markers: No specific delimiters or boundary markers are used to isolate the user-provided content from the script logic.
Capability inventory: The generated scripts possess capabilities to write to the local file system (write_env) and interact with remote services via the GitHub CLI (set_secret, set_var).
Sanitization: While the template uses printf -v for variable assignment, the subsequent storage of these values in .env files via string concatenation lacks comprehensive escaping for complex shell-special characters.
[DYNAMIC_EXECUTION]: The skill's core function is the generation of executable bash scripts from a template. It explicitly instructs the agent to create these scripts and modify their permissions using chmod +x.
The scripts are generated based on a provided library (template.sh), which limits the scope of the generated code to a predefined set of helper functions, though the final script is intended to be executed in the user's environment.

wizard