The Agent Skills Directory

[COMMAND_EXECUTION]: The skill runs local Python scripts (dockerfile_analyzer.py, compose_validator.py) to analyze Docker configurations. These operations are intended for static analysis and are contained within the skill environment.
[DATA_EXFILTRATION]: Placeholder secrets (e.g., 'password123') are present in demo configurations within the script files. These are non-sensitive examples used for testing the logic of the analyzers.
[PROMPT_INJECTION]: The skill processes user-supplied Dockerfiles and Compose files. This ingestion of untrusted data (Ingestion point: SKILL.md) without explicit boundary markers or sanitization creates a surface for indirect prompt injection, though the use of deterministic scripts for analysis (Capability inventory: scripts/dockerfile_analyzer.py) significantly mitigates this risk.

docker-development