ReasoningBank-MCP
by hanw39·★ 9·Score 38
A production-ready MCP server implementing ReasoningBank: Memory as Test-Time Compute Scaling for AI agents to evolve reasoning capabilities through experience.
Overview
ReasoningBank MCP Server is a sophisticated implementation of the 'ReasoningBank: Scaling Agent Self-Evolving with Reasoning Memory' paper. It provides AI agents with memory management capabilities through MCP, enabling them to learn from past experiences and improve over time. The system supports memory extraction, intelligent retrieval with various strategies, and advanced features like deduplication and merging to optimize memory storage and quality.
Try asking AI
After installing, here are 3 things you can ask your AI assistant:
When to choose this
Choose this server for AI agents that need to learn from experience and improve their reasoning over time, especially in multi-agent systems where specialized roles need isolated memory domains.
When NOT to choose this
Don't choose this if you need immediate results from every interaction without memory retention, or if you're working with extremely simple tasks that don't benefit from accumulated experience.
Tools this server exposes
2 tools extracted from the READMEretrieve_memory检索相关的历史经验记忆,帮助指导当前任务的执行
extract_memory从任务轨迹中提取推理经验并保存到记忆库
Comparable tools
Installation
Installation
- Clone and navigate to the repository:
git clone https://github.com/hanw39/ReasoningBank-MCP.git
cd ReasoningBank-MCP- Install dependencies:
pip install -e .Claude Desktop Configuration
Add to your Claude Desktop configuration:
{
"mcpServers": {
"reasoning-bank": {
"command": "reasoning-bank-mcp",
"env": {
"DASHSCOPE_API_KEY": "your_dashscope_api_key"
}
}
}
}Or for SSE mode:
{
"mcpServers": {
"reasoning-bank": {
"url": "http://127.0.0.1:8000/sse"
}
}
}Compare ReasoningBank-MCP with
Last updated · Auto-generated from public README + GitHub signals.