The Agent Skills Directory

[SAFE]: The skill implements a robust educational workflow without any identified security vulnerabilities. It processes technical content to generate a learning graph but does not execute external code or interact with the network. All file operations are restricted to creating a specific markdown artifact for user reference.
[INDIRECT_PROMPT_INJECTION]: The skill ingests data from local codebase modules and specifications to map atomic concepts. Although these files represent an external data ingestion surface, the skill uses them solely for conceptual mapping within a constrained markdown output format. The absence of network operations or dynamic code execution, combined with mandatory user interaction, mitigates the risk of this attack surface.

graph-learn