The Agent Skills Directory

[COMMAND_EXECUTION]: The skill's primary function is to generate a bash script based on a provided template (template.sh) that the user is instructed to execute locally. This script facilitates interactive setup processes, including opening URLs and capturing user input for configuration.
[INDIRECT_PROMPT_INJECTION]: The agent is instructed to read various repository files (e.g., README, .env.example, CI/CD workflows) to determine the necessary configuration steps for the wizard. These external files represent a potential injection surface where malicious content could attempt to influence the logic of the generated script.
Ingestion points: Reads project files like README, .env, docker-compose.yaml, and GitHub workflow files in the SKILL.md Process step 1.
Boundary markers: Not explicitly defined for the file reading process.
Capability inventory: The skill generates bash scripts capable of local file writes (.env) and executing the GitHub CLI (gh) to set repository secrets.
Sanitization: The skill does not specify sanitization for data extracted from project files, though the generated output is a script intended for human review and execution.

wizard