Graph Memory
Bound the context. Keep the memory.
A native DeepSeek Harness memory plugin that keeps recent conversation turns, archives older history, and recalls exact source-backed knowledge when it matters.
中文 ·
dsh.so ·
20-turn benchmark ·
Upgrade guide
The problem it solves
Graph Memory owns the model-visible historical surface without deleting DSH’s event log. By default it keeps the newest five completed user turns, removes completed reasoning/tool traces from future requests, and recalls relevant older or cross-session source Q/A automatically.
The 1.6 turn-memory navigation upgrade
| Before |
Now |
| Extract TASK / SKILL / EVENT directly from messages |
Create one self-contained turn summary, then derive SPO from that same sentence |
| Graph nodes could become the factual payload |
Summary, SPO, and communities only navigate; original question and final answer remain the evidence |
| Old memories from the active session could be filtered wholesale |
Exclude only sources still visible in the fresh window; archived same-session and cross-session recall share one path |
| Community expansion could pull a whole neighborhood |
Local LPA narrows candidates, query-time PPR ranks them, and only matched Q/A is recovered |
| DSH retained complete tool and reasoning traces |
Completed turns retain question + final answer; older prefixes collapse to one fixed marker |
Writing one completed turn costs exactly one auxiliary LLM call. Community detection and PPR are local. There are no hard-coded node/edge counts, semantic direction gates, or JSON repair that turns invalid output into accepted data. Read the complete design, source map, and porting sequence →
Measured first
| Real 20-turn GLM-5.2 run |
Historical native DSH baseline |
Latest Graph Memory |
Change |
| T20 first request |
56,998 tokens |
11,008 tokens |
−80.69% |
| T20 model-visible messages |
171 |
21 |
−87.72% |
| T01–T20 first-request context |
532,451 tokens |
165,896 tokens |
−68.84% |
| All measured tokens¹ |
2,487,776 |
2,327,728 |
−6.43% |
¹ The latest candidate includes 166 main-agent requests, 20 turn extractions, and 41 embedding requests; the historical baseline made 77 main requests. DSH commits and nondeterministic tool loops differ, so this is not a simultaneous strict A/B. First-request context is the direct takeover metric; the full bill remains visible.
20/20 scenario turns passed · 20/20 structured extractions succeeded · 0 quarantined · 20 turn summaries · 92 SPO triples · 30 communities · 20 summary vectors. T11, T19, and T20 automatically recalled out-of-window memory with exact source question and final answer.
Read the Markdown benchmark, per-turn data, method, and limits →
Memory survives the context window
The graph is a navigation layer, not a replacement for evidence. TASK, SKILL, and EVENT nodes point back to the original user question and final visible answer; recalled context includes those exact source messages.
Install on DeepSeek Harness
1.6.0-beta.17 fixes producer attribution on DSH 0.1.7+ / Session V4 (#113), extraction on models that cannot use off (#117), and configurable SQLite lock waiting. Context-window and recall policies remain unchanged. Fix details and recovery instructions (Chinese) →
Node.js 22.13+ · no DSH fork · until npm 1.6 is published, install the pinned GitHub release:
npx @deepseek-ai/dsh plugin --profile web add github:adoresever/graph-memory#v1.6.0-beta.17
npx @deepseek-ai/dsh --profile web --dump-config
npx @deepseek-ai/dsh web
The npm registry still serves the old 1.5.8; do not use it to validate DSH. Switch to npx @deepseek-ai/dsh plugin --profile web add graph-memory only after npm view graph-memory version reports 1.6.0-beta.17 or newer.
Confirm that graph-memory/dsh is active under Settings → Plugins. The default database is $DSH_HOME/graph-memory/graph-memory.db, normally ~/.dsh/graph-memory/graph-memory.db.
What ships
| Capability |
Implementation |
| Context takeover |
Configurable newest-N completed turns; one archive marker replaces the older model surface |
| Lightweight extraction |
Only the user question and final answer; strict structured tool contract; no reasoning/tool transcript ingestion |
| Query-first recall |
Vector Top-K with FTS5 fallback; exact source Q/A travels with graph hits |
| Durable memory |
Local SQLite, stable provenance, cross-session and cross-project recall |
| Failure behavior |
Invalid extraction is quarantined; foreground conversation continues; bad data is not repaired or persisted |
| Host support |
Native DSH/Cordis adapter; maintained OpenClaw Context Engine adapter |
Optional embeddings
Graph Memory supports OpenAI-compatible embedding endpoints. Without embeddings it falls back to FTS5 and does not block conversation.
export GRAPH_MEMORY_EMBEDDING_API_KEY='replace-with-your-key'
export GRAPH_MEMORY_EMBEDDING_BASE_URL='https://dashscope.aliyuncs.com/compatible-mode/v1'
export GRAPH_MEMORY_EMBEDDING_MODEL='text-embedding-v4'
export GRAPH_MEMORY_EMBEDDING_DIMENSIONS='1024'
dsh web
DSH tools and extraction route
| Tool |
Purpose |
gm_status |
Store, extraction, recall, vector, and retention state |
gm_search |
Explicit graph-memory search |
gm_record |
Deterministically persist a TASK, SKILL, or EVENT |
gm_stats |
Graph and retention receipts |
gm_maintain |
One bounded maintenance tick |
gm_retry_extraction |
Explicitly retry pending or quarantined extraction |
Automatic recall needs no tool call. Extraction may use a dedicated model via GRAPH_MEMORY_LLM_PROVIDER and GRAPH_MEMORY_LLM_MODEL; optional reasoning and output controls are GRAPH_MEMORY_LLM_REASONING_EFFORT and GRAPH_MEMORY_LLM_MAX_TOKENS.
Without an explicit effort, extraction reads the exact model capability: prefer off, otherwise the host’s first advertised effort, or omit the field on non-reasoning models. Route/provider failures keep source Q/A pending and report the cause in logs and gm_status. To recover old quarantined records, enable assistantTools: all and invoke gm_retry_extraction. Configure SQLite lock waiting with GRAPH_MEMORY_DB_BUSY_TIMEOUT_MS (default: 5000 milliseconds).
OpenClaw compatibility
openclaw plugins install graph-memory
openclaw plugins enable graph-memory
openclaw gateway restart
Activate the Context Engine slot in ~/.openclaw/openclaw.json:
{
"plugins": {
"slots": { "contextEngine": "graph-memory" },
"entries": { "graph-memory": { "enabled": true } }
}
}
Graph Memory Pro
The repository also contains an experimental read-only DSH Pro Lite Host + Client plugin backed by Community SQLite. The 2D/3D graph workbench, split view, and controlled drag-to-context remain planned. See dsh-pro/README_CN.md.
Verification and limits
Current beta 1.6.0-beta.17 passes 149/149 automated tests, including DSH 0.2.0-rc.2’s official V4 validator and real SQLite write contention. The GLM-5.2 20-turn figures above come from the navigation-upgrade run on 2026-09-09; this compatibility patch did not rerun that paid experiment.
- Structured extraction still depends on model contract compliance: the historical run succeeded 20/20 times. Contract failures stay quarantined; route/provider failures remain pending for explicit retry after configuration is corrected.
- Recall is bounded by configurable Top-K. Focused probes succeeded; one broad multi-topic query can require a larger Top-K or separate questions.
- The published run is an engineering workload, not a universal LoCoMo/LongMemEval score.
- The design, source-code map, and porting sequence for the summary + SPO navigation + exact-Q/A upgrade are documented in the Chinese upgrade guide.
Reproduce it from benchmarks/dsh-context-takeover/. Raw conversations, provider responses, local paths, and credentials are excluded.
Development
npm install
npm test
npm run build
npm run verify:package
MIT © 2026 adoresever · Asset and trademark notes
评论