OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
Agent Memory & RAG

Dynamics Memory

Dynamics Memory is a developer tool that converts LLM-agent project interactions into a bounded, updateable long-term memory store, with source-linked evidence, recall, feedback, and human review for conflicts.

Official resource ↗Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Dynamics Memory — OpenViST project profile in Agent Memory & RAG.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use it when an LLM agent must carry project context across sessions and its memory needs bounded growth, source traceability, and a review path for conflicts.

Who it is for

  • Developers building LLM agents that need to resume project work across sessions.
  • Teams evaluating memory-update behavior or handling contradictory project interaction history.

Documented uses

  • Retain useful project state without placing complete interaction logs into every prompt.
  • Extract and select candidate memories from ongoing interaction evidence.
  • Recall relevant memories in a later session and apply feedback to memory value.
  • Route conflicting memory proposals for human review.
  • Assess memory behavior changes with causal-replay evaluations.
Before you choose

Requirements & limits

  • The core memory engine is described as LLM-free, but semantic judgment and relevant-set operations use a separate provider path; setup lists a ZAI_API_KEY.
  • The repository describes an OpenCode plugin and sidecar integration, with SQLite-backed evidence and task stores.
  • The memory pools have stated capacity limits, while unresolved conflicts are routed to a human-review queue.
  • The README labels the project a research preview and identifies PENDING-08 as awaiting manual acceptance.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Start the memory sidecar for a project

Install the listed Python dependencies, configure the environment file, and start the project-scoped memory server.

Before you start

  • Python environment
  • NumPy and pytest installed with pip
  • A copy of the repository
  • A .env file populated with ZAI_API_KEY

Steps

  1. Run the command below.

    Documented command
    pip install numpy pytest

  2. Copy .env.example to .env and set ZAI_API_KEY.

  3. Run the command below, replacing `` with the project directory.

    Documented command
    python -m hybrid_memory.server --project

Expected outcome

The memory server starts with the specified project directory.

Guide limitations

  • This setup path does not document an end-to-end agent session or retrieval workflow.
  • The README labels the project a research preview and notes PENDING-08 as awaiting manual acceptance.
Documentation 1 ↗
Source-based practical guide

Load the bridge through OpenCode

Use the repository’s OpenCode plugin to load the memory bridge and launch the sidecar.

Before you start

  • Node package manager
  • OpenCode CLI installed as opencode-ai
  • A local copy of the Dynamics Memory repository

Steps

  1. Install the CLI with the command below.

    Documented command
    npm i -g opencode-ai

  2. Change into the repository directory.

  3. Run opencode; the README says the .opencode/plugin/memory-bridge.ts plugin loads and launches the sidecar.

Expected outcome

The repository’s OpenCode plugin bridge is loaded and starts the sidecar.

Guide limitations

  • For use in another project, the README describes linking memory-bridge.ts into that project’s .opencode/plugin/ directory and resolving @opencode-ai/plugin from the link’s real path.
  • The README does not provide a full configuration or verification procedure for a separate project.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official resource ↗Official repository ↗
Source record details
Source availability
Active
Sources checked
16 October 2026
Original website title
1173591564/Dynamics-memory · Official homepage title · retained in its source language
How we review sourcesSuggest a correction