The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill instructions require the agent to read various repository files to determine the necessary setup steps and variables.
Ingestion points: The agent is instructed to read .env, .env.example, README, docker-compose*, framework configurations, and GitHub workflow files (SKILL.md).
Boundary markers: There are no explicit instructions or delimiters to help the agent distinguish between legitimate configuration data and potential malicious instructions embedded within these files.
Capability inventory: The generated wizard script has the capability to modify local environment files via write_env and upload secrets to GitHub Actions via set_secret using the gh CLI (template.sh).
Sanitization: The skill does not define specific sanitization or validation logic for the data extracted from the repository files before using it to generate the wizard script.
[COMMAND_EXECUTION]: The primary purpose of the skill is to author and encourage the execution of bash scripts on the user's local machine.
Evidence: The skill instructions guide the agent to create a script based on template.sh, apply chmod +x, and instruct the user on how to run it. While the script is intended to be transparent and manual, the generation process itself is a vector for delivering executable commands to the user's environment.

wizard