你的AI失忆了。我们修复了它。

每个LLM都是从零开始的每一次对话。Stash为你的代理提供持久记忆——它能在会话间记忆、回忆、整合和学习。不再从零开始解释自己。

开源。自架。支持任何兼容MCP的代理。

---

不想自己托管？ usestash.io 是托管的云端版本——用 Google 登录，复制一个 MCP URL，就完成了。免费开始。

---

快速启动

仅此而已。Postgres + pgvector、迁移、带后台整合的 MCP 服务器——全部集中在一个命令里。

** 下一页：** 入门指南 — 连接你的MCP客户端，运行init / remember / recall]，确认一切正常。

完全本地（无云API）: Ollama 设置指南 — host Ollama + Docker Compose，私有嵌入和 reasoner。

阿特拉斯云

Atlas Cloud 是一个全模态 AI 推理平台，为开发者提供单一的 AI API 访问视频生成、图像生成和大型语言模型 API。你无需管理多个供应商集成，只需连接一次，即可统一访问涵盖所有模态的 300+ 个精心策划的模型。

查看Atlas Cloud的编码计划促销：https://www.

Stash

Your AI has amnesia. We fixed it.

Every LLM starts every conversation from zero. Stash gives your agent persistent memory — it remembers, recalls, consolidates, and learns across sessions. No more explaining yourself from scratch.

Open source. Self-hosted. Works with any MCP-compatible agent.

---

Don’t want to self-host? usestash.io is the hosted cloud version — sign in with Google, copy one MCP URL, and you’re done. Free to start.

---

Quick Start

git clone https://github.com/alash3al/stash.git
cd stash
cp .env.example .env   # edit with your API key + model
docker compose up

That’s it. Postgres + pgvector, migrations, MCP server with background consolidation — all in one command.

Next: Getting Started guide — connect your MCP client, run init / remember / recall, and verify everything works.

Fully local (no cloud API): Ollama setup guide — host Ollama + Docker Compose, private embeddings and reasoner.

Atlas Cloud

Atlas Cloud is a full-modal AI inference platform that gives developers a single AI API to access video generation, image generation, and LLM APIs. Instead of managing multiple vendor integrations, you connect once and get unified access to 300+ curated models across all modalities.

Check out Atlas Cloud’s coding plan promotion: https://www.atlascloud.ai/console/coding-plan

Stash already supports Atlas Cloud through its OpenAI-compatible API. Set your .env like this:

STASH_OPENAI_API_KEY=your-atlas-cloud-api-key
STASH_OPENAI_BASE_URL=https://api.atlascloud.ai/v1
STASH_EMBEDDING_MODEL=text-embedding-3-small
STASH_REASONER_MODEL=deepseek-ai/DeepSeek-V3-0324
STASH_VECTOR_DIM=1536

Atlas Cloud works well here because Stash only needs:

• an embeddings model for vectorization
• a chat-capable reasoning model for consolidation
• an OpenAI-compatible base URL and API key

See Getting Started for a fuller configuration checklist.

MCP Client Setup

After docker compose up, Stash exposes an MCP server over SSE at:

http://localhost:8080/sse

Point any MCP-compatible client at that URL. Example configs:

Cursor — ~/.cursor/mcp.json

{
  "mcpServers": {
    "stash": {
      "url": "http://localhost:8080/sse"
    }
  }
}

Claude Desktop — claude_desktop_config.json

{
  "mcpServers": {
    "stash": {
      "url": "http://localhost:8080/sse"
    }
  }
}

OpenCode — ~/.config/opencode/config.json

{
  "mcp": {
    "stash": {
      "type": "remote",
      "url": "http://localhost:8080/sse",
      "enabled": true
    }
  }
}

Windsurf — ~/.codeium/windsurf/mcp_config.json

{
  "mcpServers": {
    "stash": {
      "url": "http://localhost:8080/sse"
    }
  }