mirror of
https://github.com/langchain-ai/langgraph.git
synced 2026-08-22 07:32:25 +02:00
Warnings in _check_allowed and _check_allowed_method now track already-warned types/methods per ext_hook closure, so the same warning is only emitted once. Added register_safe_types() to _msgpack.py as a public API for users to add custom types to the global safe-types set, exported from langgraph.checkpoint.serde. Co-authored-by: Sydney Runkle <54324534+sydney-runkle@users.noreply.github.com>