The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill's implementation playbook and scripts download automation tools and dependencies from well-known registries including npm, PyPI, and the GitHub Actions marketplace (e.g., official actions from AWS, Hashicorp, and Snyk).
[COMMAND_EXECUTION]: The provided development environment setup script (scripts/setup-dev-environment.sh) executes multiple shell commands to install prerequisites via npm and pip, manage Docker services, and initialize local databases.
[PROMPT_INJECTION]: The skill processes untrusted local data which creates a surface for indirect prompt injection.
Ingestion points: The WorkflowAnalyzer class in resources/implementation-playbook.md reads project-specific configuration files (.github/workflows/*.yml, .gitlab-ci.yml, Jenkinsfile) and documentation files (README.md, README.rst, README.txt).
Boundary markers: No specific delimiters or instructions are used to isolate the parsed file content from the agent's core instructions.
Capability inventory: The agent has the ability to generate and write workflow files, configure CI/CD secrets, and execute infrastructure management commands via the provided templates.
Sanitization: While it uses yaml.safe_load for Python parsing, it does not validate or sanitize the natural language content of documentation files to prevent embedded instructions from influencing agent behavior.

cicd-automation-workflow-automate