MemOS Long-Term Memory & Cross-Session Graph Store

@mem-os/vector-store · v2.4.0

A persistent cognitive memory and cross-session graph store for DeepSeek Harness. Combines working memory, episodic traces, and semantic networks with automated forgetting curves.

Long-Term-MemoryEpisodic-StoreVector-SearchCognition
GitHub Stars
67 K+
+12.4% this month
Monthly Downloads
298 K+
Monthly registry pulls
Reach Score
99.4/ 100
Top Tier Ecosystem
Runtime
Cordis v3+
Node 18+ / Bun / Deno

Installation & Integration

CLI one-click launch, package managers, and Cordis integration

bash
manager:
$ pnpm add @mem-os/vector-store

Architecture

MemOS solves agent amnesia across ephemeral sessions. Standard LLM agents start completely tabula rasa each run, forgetting developer coding styles, architectural patterns, and previous debugging breakthroughs. This plugin implements a tri-tier cognitive memory architecture inside Cordis. Transient working memory maintains active turns, episodic memory logs chronological triumphs and mistakes, and semantic graphs consolidate recurring truths. Automated Ebbinghaus forgetting curves prune stale noise, allowing the agent to continuously compound domain expertise.

Architectural Principles & Constraints

01
Strict Type Isolation

Guaranteed by TypeScript compile-time contracts, inter-plugin event bus calls enjoy zero-drift safety.

02
Sub-Millisecond Hot Reload

Supports dynamic runtime mounting and graceful unloading without restarting the primary host process.

03
Deterministic State Machine

Embeds multi-phase execution lifecycle guards, preventing context loss during long-horizon reasoning.

04
Zero Native Build Dependencies

Designed for lightweight cross-platform environments, booting instantly across Node.js, Bun, and Deno.

Core Features

01
Tri-Tier Cognitive Model: Working context, episodic event logs, and semantic concepts in sync
02
Ebbinghaus Decay Pruning: Fades out stale one-off instructions while reinforcing habitual preferences
03
Multi-Tenant Partitioning: Strictly isolates memory profiles by user ID, workspace, and team
04
Embedded Zero-Config Storage: Runs natively on local SQLite-Vector with zero external database setup

Core Workflow

01

Experience Distillation

Extracts key facts, user preferences, and lesson summaries upon task completion.

02

Hierarchical Ingestion

Classifies memories into chronological episodes or interconnected concepts.

03

Contextual Semantic Recall

Dynamically retrieves salient past memories matching current prompts.

04

Forgetting Curve Decay

Applies time-decay algorithms to prune trivial logs while cementing key insights.

Configuration Parameters Reference (YAML / JSON)

ParameterTypeDefaultDescription
decayRatenumber0.05Time decay factor for memory retention
maxRecallChunksnumber3Maximum salient memory chunks recalled
storagePathstring./.dsh/memories.dbFilepath to local SQLite memory database

Use Cases

Enterprise Production Agent

Relies on microkernel lifecycle guards and fault-tolerant state machines for continuous reliability.

SWE-bench Benchmark Evaluation

Native integration with SWE-bench workflows, automatically capturing diffs and verification metrics.

Autonomous Code Refactoring

Separates reasoning from tool actions to independently locate and refactor multi-file codebases.

Cross-Tool Workflow Automation

Safely orchestrates events across sandboxes to seamlessly link enterprise developer tooling.

Best Practices

01
Sandbox Permission Guard

Strictly isolate sub-process calls and network scope; deploy within Docker containers in production.

02
Exponential Backoff Retries

Configure adaptive exponential retries with strict timeouts to mitigate upstream model rate limits.

03
Session State Checkpointing

Persist state machine snapshots to survive hardware interruptions and resume instantly without loss.

04
Full Trajectory Audit Logs

Enable full trace logging, aggregating reasoning thought streams and tool I/O into your observability hub.

FAQ

Q1:How to handle timeouts in long-running autonomous tasks?

Increase the timeout parameter inside your YAML configuration and dispatch periodic heartbeat signals across the Cordis event bus. For long-running tool execution and model reasoning, configure persistent session snapshotting so suspended tasks can safely resume their exact context after interruptions, preventing the kernel from recycling active agent sessions prematurely.

Q2:How to capture and stream the model reasoning thought process?

The harness runtime natively provides end-to-end streaming hooks while its state machine automatically intercepts and strips <think> reasoning tags from model responses. Subscribe directly to the onThink event listener to consume live reasoning token streams in real-time, delivering typewriter animation to the user interface while persisting full trajectories for audit compliance.

Q3:How to resolve dependency conflicts across multiple plugins?

Cordis microkernel uses directed acyclic graph topological sorting to dynamically resolve plugin dependencies. When shared services or runtime versions conflict, assign distinct isolated namespaces at the application entrypoint. Leveraging context injection alongside lazy activation ensures dependencies load strictly on-demand when tools are triggered, maintaining system stability and preventing memory bloat.

Q4:How to enforce permissions and sandbox isolation in production?

Combine isolated container sandboxing with fine-grained capability checks to prevent plugins from accessing sensitive files or unauthorized external networks. Enforce explicit runtime system call whitelisting through the microkernel, executing all third-party tool scripts inside isolated ephemeral containers to block arbitrary code execution and eliminate security privilege escalation risks entirely.

Recommended Ecosystem Plugins

Explore related Cordis extensions designed to work synergistically