The Agent Skills Directory

[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface by ingesting data from the local repository to guide its configuration scoping.
Ingestion points: The agent is instructed to read various project files including .env.example, README, and GitHub workflow files to determine required variables (SKILL.md).
Boundary markers: There are no explicit instructions to use delimiters or ignore instructions embedded within the processed files during the scanning phase.
Capability inventory: The generated bash scripts have the capability to write to local files, open browser URLs, and update GitHub repository secrets using the gh tool (template.sh).
Sanitization: The skill does not describe specific sanitization or validation logic for the information extracted from project files before it influences the generated script.
[COMMAND_EXECUTION]: The provided template.sh library defines functions that execute shell commands to interact with the environment.
It utilizes platform-specific commands like xdg-open, open, or explorer.exe to launch URLs in the user's browser.
It leverages the GitHub CLI (gh) for managing repository secrets and variables, assuming the tool is installed and authenticated.

wizard