The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill documents a workflow for aggregating and resolving external dependency files which could potentially contain malicious instructions if sourced from untrusted third-party repositories.
Ingestion points: The skill reads requirements.txt and pyproject.toml files from core and custom node repositories (SKILL.md).
Boundary markers: No boundary markers or specific delimiters are used when concatenating these external files into declared.txt.
Capability inventory: The skill utilizes git clone to fetch external content, cat to read files, and uv pip compile to resolve dependencies.
Sanitization: No sanitization of the requirement file content is performed before resolution, as the process relies on the underlying package manager's validation.

comfy-build-pins