Implementation based on the paper "ReasoningBank: Scaling Agent Self-Evolving with Reasoning Memory"
{
"mcpServers": {
"reasoningbank-mcp": {
"command": "<see-readme>",
"args": []
}
}
}No install config available. Check the server's README for setup instructions.
Are you the author?
Add this badge to your README to show your security score and help users find safe servers.
Implementation based on the paper "ReasoningBank: Scaling Agent Self-Evolving with Reasoning Memory"
Is it safe?
No package registry to scan.
No authentication — any process on your machine can connect.
MIT. View license →
Is it maintained?
Last commit 160 days ago. 9 stars.
Will it work with my client?
Transport: stdio. Works with Claude Desktop, Cursor, Claude Code, and most MCP clients.
No automated test available for this server. Check the GitHub README for setup instructions.
No known vulnerabilities.
This server is missing a description. Tools and install config are also missing.If you've used it, help the community.
Add informationHave you used this server?
Share your experience — it helps other developers decide.
Sign in to write a review.
Dynamic problem-solving through sequential thought chains
A Model Context Protocol server for searching and analyzing arXiv papers
An open-source AI agent that brings the power of Gemini directly into your terminal.
The official Python SDK for Model Context Protocol servers and clients
MCP Security Weekly
Get CVE alerts and security updates for ReasoningBank MCP and similar servers.
Start a conversation
Ask a question, share a tip, or report an issue.
Sign in to join the discussion.
随着大语言模型代理在持久性现实角色中的日益普及,它们自然会遇到连续的任务流。然而,一个关键的限制是它们无法从累积的交互历史中学习,迫使它们丢弃宝贵的见解并重复过去的错误。基于论文ReasoningBank: Scaling Agent Self-Evolving with Reasoning Memory,我们实现了这个记忆增强推理系统,通过 MCP (Model Context Protocol) 协议为 AI 代理提供经验记忆管理能力。
ReasoningBank 提出了一种新颖的记忆框架,能够从代理自身判断的成功和失败经验中提炼出可泛化的推理策略。在测试时,代理从 ReasoningBank 中检索相关记忆来指导其交互,然后将新学到的知识整合回去,使其能够随着时间的推移变得更加强大。这种内存驱动的经验扩展为代理创建了一个新的扩展维度,使它们能够自我进化并产生新兴行为。
reasoning-bank-mcp/
├── src/
│ ├── server.py # MCP 服务器入口
│ ├── config.py # 配置管理
│ ├── tools/ # MCP 工具
│ │ ├── retrieve_memory.py # 检索记忆
│ │ └── extract_memory.py # 提取记忆
│ ├── retrieval/ # 检索策略
│ │ ├── base.py # 抽象接口
│ │ ├── factory.py # 策略工厂
│ │ └── strategies/ # 具体策略实现
│ ├── deduplication/ # 去重策略(v0.2.0+)
│ │ ├── base.py # 抽象接口
│ │ ├── factory.py # 策略工厂
│ │ └── strategies/
│ │ ├── hash_dedup.py # 哈希去重
│ │ └── semantic_dedup.py # 语义去重
│ ├── merge/ # 合并策略(v0.2.0+)
│ │ ├── base.py # 抽象接口
│ │ ├── factory.py # 策略工厂
│ │ └── strategies/
│ │ ├── llm_merge.py # LLM智能合并
│ │ └── voting_merge.py # 投票选择
│ ├── services/ # 服务层(v0.2.0+)
│ │ └── memory_manager.py # 记忆管理服务
│ ├── storage/ # 存储后端
│ │ ├── base.py # 抽象接口
│ │ └── backends/ # 具体存储实现
│ ├── llm/ # LLM 客户端
│ │ ├── base.py # 抽象接口
│ │ ├── factory.py # Provider 工厂
│ │ └── providers/ # 具体 Provider 实现
│ ├── prompts/ # 提示词模板
│ └── utils/ # 工具函数
└── data/ # 数据存储目录
├── memories.json # 记忆数据库
├── archived_memories.json # 归档记忆(v0.2.0+)
└── embeddings.json # 嵌入向量
git clone https://github.com/hanw39/ReasoningBank-MCP.git
cd ReasoningBank-MCP
pip install -e .
{
"mcpServers": {
"reasoning-bank": {
"command": "reasoning-bank-mcp",
"env": {
"DASHSCOPE_API_KEY": "你的百炼APIKEY"
}
}
}
}
1) 启动服务器:
# 使用默认配置 (127.0.0.1:8000)
python3 -m src.server --transport sse
# 或指定主机和端口
python3 -m src.server --transport sse --host 0
... [View full README on GitHub](https://github.com/hanw39/ReasoningBank-MCP#readme)