The Agent Skills Directory

[DYNAMIC_EXECUTION]: The entry point script scripts/run.py uses os.execv() to replace the current process with a Python interpreter located at a computed path (.runners-venv/bin/python3). While this is a standard method for ensuring a script runs within its intended virtual environment, process replacement using paths computed at runtime is a dynamic execution pattern.
[INDIRECT_PROMPT_INJECTION]: The skill accepts user-provided text through flags such as --title, --creator, and --subtitle. These values are directly interpolated into messages sent to sub-agents for image generation.
Ingestion points: Untrusted user input enters the workflow via metadata flags defined in SKILL.md.
Boundary markers: The pipeline instructions in SKILL.md (Step 5) suggest using double quotes to delimit these values in the prompt, which provides a basic but bypassable boundary.
Capability inventory: The skill can execute shell commands through the Bash tool and utilizes process replacement in scripts/run.py.
Sanitization: There is no evidence of explicit validation, sanitization, or filtering of the user-provided strings before they are interpolated into the prompt.

cover-maker