Window Buffer Memory
Window Buffer Memory
Section titled “Window Buffer Memory”Keeps only the last N messages in memory
What it does
Section titled “What it does”Maintains conversation history for the duration of the current execution, keeping only the last k human/AI message exchanges. Older messages are dropped from context. Unlike Local Memory or Persistent Chat Memory, this memory is not persisted across runs.
When to use it
Section titled “When to use it”Connect this to AI agents that need short-term conversational context within a single run, while bounding how much history is kept in the prompt.
Inputs and settings
Section titled “Inputs and settings”| Setting | Notes |
|---|---|
| Window Size (k) | Only the last k message exchanges are kept in memory. Number of most recent human/AI message pairs to keep; older messages are dropped from context. Default: 5. |
Outputs
Section titled “Outputs”Returns a memory dependency for conversational AI workflows.
Dependencies and credentials
Section titled “Dependencies and credentials”- No explicit credential dependency is declared in the node description.
Example workflow
Section titled “Example workflow”Connect Window Buffer Memory to an AI agent or dependency input that accepts this dependency type, set the Window Size (k), then run the agent with data from previous nodes.
Troubleshooting
Section titled “Troubleshooting”- Check that Window Size (k) reflects how much recent context you want retained — a small k can drop information the agent still needs.
- Remember this memory does not persist across separate workflow runs; use Persistent Chat Memory if you need that.
- No dedicated source test was found next to this node; verify behavior manually when changing this page.