LangChain MongoDB provides integrations between MongoDB, Atlas, LangChain, and LangGraph. Prior to langgraph-checkpoint-mongodb 0.3.0 and langgraph-store-mongodb 0.4.0, MongoDBSaver.list(), MongoDBSaver.alist(), and MongoDBStore.search() incorporate filter dictionaries into MongoDB queries without recursively rejecting keys prefixed with $. An authenticated caller who controls a filter argument through HTTP query parameters, request body fields, or agent tool arguments can inject MongoDB Query Language operators such as $regex or $where. In a multi-tenant deployment that uses the filter to enforce per-user or per-tenant isolation, injected operators can bypass intended equality filtering and expose other tenants' checkpoint or store data. Filters constructed entirely from trusted server-side values have lower practical risk. This issue is fixed in langgraph-checkpoint-mongodb 0.3.0 and langgraph-store-mongodb 0.4.0.
Summary
AIMongoDBSaver.list(), MongoDBSaver.alist(), and MongoDBStore.search() do not recursively reject MongoDB operator keys in filter dictionaries. An authenticated caller who controls a filter through HTTP parameters, request fields, or agent tool arguments can inject query operators and, where filters enforce tenant isolation, potentially access other tenants’ checkpoint or store data. No user interaction is required.
Who should check
Check deployments using PyPI langgraph-checkpoint-mongodb versions below 0.3.0 or langgraph-store-mongodb versions below 0.4.0.
What to do
- 1Upgrade langgraph-checkpoint-mongodb to 0.3.0 or later.
- 2Upgrade langgraph-store-mongodb to 0.4.0 or later.