The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests data from untrusted local project metadata files (such as pyproject.toml, package.json, Cargo.toml, and go.mod) to automate repository configuration.
Ingestion points: Project metadata is read from files in the repository root to extract package names, versions, licenses, and descriptions (documented in CICD.md).
Boundary markers: The instructions do not specify any delimiters or boundary markers to distinguish extracted data from instructions during processing.
Capability inventory: The skill executes shell commands via git and gh (initialization, configuration, cloning, committing, and pushing) and performs file system writes to create CI/CD workflows, licenses, and README files.
Sanitization: There are no explicit instructions to sanitize or validate the extracted metadata before it is interpolated into shell commands (e.g., gh repo edit --description) or written to documentation files.
[COMMAND_EXECUTION]: The skill makes extensive use of the git and github-cli (gh) tools to perform repository management tasks.
The skill runs git init, git config, git clone, git commit, and git push to manage local and remote repositories.
It uses gh repo create, gh repo edit, and gh api to interact with the GitHub platform.
These actions are consistent with the skill's primary purpose of repository initialization.

init-github-repo