Agent Memory V2 Node
Beta lossless agent memory. Persists complete session JSONL values before replacing oversized active-context content with compact, retrievable references.
:::warning This node is in beta and may not be fully tested or documented. Use with caution. ::: Beta lossless agent memory. Persists complete session JSONL values before replacing oversized active-context content with compact, retrievable references.
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 | Hard cap on the context budget in tokens. The active model window is clamped down to this value when it is larger; a smaller model window is used as-is. | 128000 | |
| Model Context Window % | modelContextWindowPercent | botdojo/types/number | Fraction of the active model context window available to memory | 0.9 | ✅ |
| Summarization Trigger % | compressPercentage | botdojo/types/number | Summarize older conversation history at this fraction of the context budget | 0.85 | ✅ |
| Post-Summarization Target % | compactionTargetPercentage | botdojo/types/number | Soft active-memory target after summarization; atomic recent tool groups remain intact | 0.3 | ✅ |
| Recent Tail % of Threshold | compactionRecentTailPercentage | botdojo/types/number | Token budget for exact recent messages, measured as a fraction of the summarization threshold | 0.2 | ✅ |
| Recent Tail Soft Limit Multiplier | compactionRecentTailSoftLimitMultiplier | botdojo/types/number | Allows a complete recent atomic group to extend beyond the target tail budget | 1.5 | ✅ |
| Minimum Recent Groups | compactionMinRecentGroups | botdojo/types/number | Minimum number of newest message/tool groups retained exactly | 2 | ✅ |
| Summary Context Limit % | compactionSummaryPercentage | botdojo/types/number | Maximum fraction of the memory context budget available to the generated checkpoint | 0.05 | ✅ |
| Summary Source Ratio | compactionSummarySourcePercentage | botdojo/types/number | Desired summary-token budget as a fraction of the historical tokens being summarized | 0.2 | ✅ |
| Summary Minimum Tokens | compactionSummaryMinTokens | botdojo/types/number | Minimum checkpoint budget before applying the context-percentage and absolute caps | 1000 | ✅ |
| Summary Maximum Tokens | compactionSummaryMaxTokens | botdojo/types/number | Absolute generated-checkpoint token ceiling | 10000 | ✅ |
| Minimum Messages to Summarize | minMessagesToCompress | botdojo/types/number | Skip summarization when fewer historical messages are eligible | 5 | ✅ |
| Retrieval Preview Window (chars) | maxDefaultWindow | botdojo/types/number | Compatibility setting retained in serialized V2 configuration | 1000 | ✅ |
| Enable Token Counting | enableTokenCounting | botdojo/types/boolean | Use tokenizer-based counts instead of the fast character estimate | false | ✅ |
| Compaction Language Model | languageModel | botdojo/types/language_model | Language model used only when active session history crosses the compaction threshold | null | ✅ |
| Compaction Prompt | compactionPrompt | botdojo/types/string | Instructions used to turn older session messages into a compact context checkpoint. Leave blank to use the built-in default prompt; provide text to override it. | ✅ | |
| Externalize At (chars) | externalizeAtChars | botdojo/types/number | Persist and immediately replace very large tool outputs with compact JSONL references | 6000 | ✅ |
| Preview (chars) | previewChars | botdojo/types/number | Maximum deterministic preview before a compact memory reference | 300 | ✅ |
| View Inline Limit (chars) | viewInlineChars | botdojo/types/number | Keep already-bounded workspace view chunks inline up to this safety limit | 15000 | ✅ |
Output
| Label | Property | Type | Description | Default Value |
|---|---|---|---|---|
| Memory | memory | botdojo/interface/memory | Standalone Agent Memory V2 instance | null |