The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes source code, which is untrusted data, and instructs the agent to perform refactoring and deletions. This creates a surface where instructions embedded in the codebase could manipulate the agent's behavior.
1. Ingestion points: Codebase files and project structures provided by the user for refactoring.
1. Boundary markers: The skill does not provide specific delimiters or instructions to ignore instructions embedded in code comments or metadata.
1. Capability inventory: The agent typically has read and write access to the file system to implement the refactoring goals.
1. Sanitization: No sanitization or validation of the codebase content is performed before processing.
[EXTERNAL_DOWNLOADS]: The skill includes links to external educational resources for software design principles.
References: Documentation links to well-known technical platforms and academic sources, including InfoQ, YouTube, and specialized development blogs (e.g., infoq.com, curtclifton.net, grugbrain.dev, peps.python.org). These links are for reference purposes and are not executed by the skill logic.

reducing-entropy