The Agent Skills Directory

[SAFE] (SAFE): No malicious patterns, prompt injections, or data exfiltration attempts were detected. The content consists of structural Python code snippets for LangGraph (TypedDict, Pydantic, reducers) and educational guidance.
[NO_CODE] (INFO): The skill contains no actual executable scripts (.py, .js, .sh). It provides markdown-based context and examples to assist the agent in designing LangGraph workflows. Use of standard libraries like typing and operator is benign.
[DATA_EXPOSURE] (SAFE): The provided code snippets demonstrate state management for standard data types and chat messages. There are no hardcoded credentials, sensitive file paths, or unauthorized network operations.

langgraph-state