The Agent Skills Directory

[SAFE]: The skill provides standard, high-quality templates for application containerization. It correctly advises excluding environment variables ('.env') from the image build context using '.dockerignore', preventing accidental credential exposure. It promotes the use of official, slim base images (Alpine) which reduces the potential attack surface.\n- [INDIRECT_PROMPT_INJECTION]: The skill ingests data from local project files to detect the application runtime, which presents a surface for instructions embedded in project metadata. This is a standard requirement for the skill's primary function and the risk is considered negligible given the specific detection task.\n
Ingestion points: The agent reads manifest files such as 'package.json', 'requirements.txt', 'go.mod', and 'Cargo.toml' from the project directory to identify the runtime environment.\n
Boundary markers: No specific delimiters or instructions to ignore embedded commands are provided for the manifest inspection phase.\n
Capability inventory: The agent is tasked with generating and writing several configuration files to the local file system (Dockerfile, docker-compose.yml, .dockerignore).\n
Sanitization: The content from project manifests is used for runtime identification only; no specific sanitization or filtering is performed on this input.

adding-docker