Agent Memory Node
warning
This node is in beta and may not be fully tested or documented. Use with caution.
DB-backed agent memory using DocumentStore + IndexStore. Reads agent context from flow headers (agent_id). Provides context-window management with compression-on-retrieval, JSONL session logging, and transcript finalization. Compressed messages include a source reference to the raw JSONL log.
Properties
| Label | Property | Type | Description | Default Value | Is Required |
|---|---|---|---|---|---|
| Memory Key | memoryKey | botdojo/types/string | agent_memory | ||
| Max Context Window (tokens) | maxContextWindow | botdojo/types/number | Maximum tokens before pruning oldest messages | 128000 | |
| Compress Percentage | compressPercentage | botdojo/types/number | Start compressing older messages at this fraction of the context window (0.0-1.0) | 0.8 | ✅ |
| Compression Window (chars) | maxDefaultWindow | botdojo/types/number | Characters preserved on each side when compressing a message (middle is truncated) | 1000 | ✅ |
| Max Tool Output (chars) | maxToolOutputChars | botdojo/types/number | Auto-compress tool outputs exceeding this character limit on ingest | 5000 | ✅ |
| Enable Token Counting | enableTokenCounting | botdojo/types/boolean | Use tiktoken for accurate token counts (false = char/4 estimate, faster but less precise) | false | ✅ |
Output
| Label | Property | Type | Description | Default Value |
|---|---|---|---|---|
| Memory | memory | botdojo/interface/memory | Agent memory instance backed by DocumentStore | null |