The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the user to execute shell commands such as git ls-files, git add, and railway up to manage the deployment process and verify project integrity.
[CREDENTIALS_SAFE]: The skill implements a security best practice by checking for sensitive files like .env, .pem, and service account keys in the Git index before allowing a commit or deployment. It correctly advises that credentials should be stored in environment files and never committed to the repository.
[INDIRECT_PROMPT_INJECTION]: The skill processes project file listings which could theoretically be manipulated by an attacker to include malicious filenames.
Ingestion points: Output from Git commands (git ls-files, git status).
Boundary markers: Instructions require the agent to display the full file list to the user for manual review and confirmation before performing actions like git add -A or committing.
Capability inventory: The skill has access to shell command execution for Git management and deployment tools.
Sanitization: File lists are filtered using specific regex patterns to identify secrets.

publicar