Cognitive Exoskeleton Server

个人认知外骨骼 — 基于知识图谱 + LLM 推理的「第二大脑」MCP Server

Your Personal Cognitive Exoskeleton — a "second brain" MCP Server powered by knowledge graph + LLM reasoning

---

中文文档

Cognitive Exoskeleton 不是普通的搜索工具。它从你的笔记中构建动态知识图谱，然后利用 LLM 推理能力主动发现知识盲点、寻找隐藏的跨领域关联、追踪你对某个概念的理解如何随时间演变，并通过碰撞不同领域的想法来激发创造性灵感。

• 所有数据完全本地存储（SQLite），隐私安全
• 零配置：默认通过 MCP Sampling 复用客户端的 LLM，也可自带 API（Hy3、OpenAI、Ollama、vLLM 等）
• 即插即用：兼容 Cursor、CodeBuddy、WorkBuddy、Cline 等主流 MCP 客户端

功能特性

提供 8 个 MCP 工具，分为四个层次：

层次	工具	功能
基础层	ingest_note	从笔记/文档中抽取实体和关系，写入知识图谱
	query_mind	基于知识图谱回答问题，支持浅层/深层检索
	recall_context	写作时自动召回相关但可能遗忘的旧笔记
推理层	discover_connections	发现不同领域间隐藏的、非显而易见的知识关联
	detect_blindspots	分析某话题的知识覆盖度，识别盲点、矛盾和缺失视角
	analyze_cognitive_topology	生成「认知画像」：知识孤岛、桥梁概念、密集区/空白区
时间层	trace_concept_evolution	追踪你对某个概念的理解如何随时间变化
灵感层	spark_serendipity	碰撞两个不同领域的概念，激发跨域创造性灵感

快速开始

环境要求：Node.js >= 18

# 克隆并安装
git clone https://github.com/Tencent-Hunyuan/Hy3.git
cd Hy3/rhinobird2026/cognitive-exoskeleton-mcp
npm install && npm run build

# 启动（零配置 — 自动复用 Cursor/WorkBuddy 的模型）
node dist/index.js

零配置模式：默认情况下，MCP Server 通过 MCP Sampling 协议复用客户端（Cursor、WorkBuddy 等）已配置的 LLM 模型，无需单独配置 API key。

如需使用独立的 LLM API，可配置环境变量切换到 Direct 模式：

export LLM_MODE=direct
export LLM_API_BASE="http://127.0.0.1:8000/v1"
export LLM_API_KEY="EMPTY"
export LLM_MODEL_NAME="hy3"
node dist/index.js

环境变量

变量	说明	默认值
LLM_MODE	LLM 调用模式：sampling（委托客户端）或 direct（直连 API）	自动检测*
LLM_API_BASE	(Direct 模式) OpenAI 兼容 API 的基础 URL	http://127.0.0.1:8000/v1
LLM_API_KEY	(Direct 模式) LLM 提供商的 API Key	EMPTY
LLM_MODEL_NAME	(Direct 模式) 使用的模型名称	gpt-4o-mini
COGNITIVE_DB_PATH	SQLite 数据库文件路径	./cognitive.db

* 自动检测逻辑：如果 LLM_API_BASE 和 LLM_API_KEY 都已配置（且 key 不是 EMPTY），则使用 direct 模式；否则使用 sampling 模式。

多模型支持（Direct 模式）

以下配置仅在 LLM_MODE=direct 时需要：

模型提供商	LLM_API_BASE	LLM_MODEL_NAME
Hy3（本地 vLLM）	http://127.0.0.1:8000/v1	hy3
Hy3（官方 API）	https://api.hunyuan.tencent.com/v1	hy3
OpenAI	https://api.openai.com/v1	gpt-4o
Ollama（本地）	http://localhost:11434/v1	qwen2.5:32b
vLLM（任意模型）	http://127.0.0.1:8000/v1	<模型名>

MCP 客户端配置

Cursor — .cursor/mcp.json（零配置，复用 Cursor 的模型）：

{
  "mcpServers": {
    "cognitive-exoskeleton": {
      "command": "node",
      "args": ["<项目路径>/dist/index.js"],
      "env": {
        "LLM_MODE": "sampling"
      }
    }
  }
}

CodeBuddy / WorkBuddy — CLI 命令：

codebuddy mcp add cognitive-exoskeleton \
  --command "node" \
  --arg "<项目路径>/dist/index.js" \
  --env LLM_MODE=sampling

Sampling 模式说明：MCP Server 通过 MCP Sampling 协议将 LLM 调用委托给客户端。客户端会使用自己已配置的模型（如 Cursor 配置的 Claude/GPT、WorkBuddy 配置的模型等），用户无需为 MCP Server 单独申请或配置 API key。每次 LLM 调用时客户端会通知用户。

使用示例

导入笔记：

用户：请把这篇笔记导入知识图谱：
"分布式系统遵循 CAP 定理，真正选择是在 CP 和 AP 之间。"
→ 自动抽取 CAP定理、一致性、可用性等实体及关系

图谱问答：

用户：我对 CAP 定理了解多少？
→ 从图谱检索相关实体，LLM 推理后返回结构化答案

写作时召回：

用户：我正在写关于数据库一致性模型的文字...
→ 召回 3 个月前关于 CAP 定理的旧笔记

发现隐藏关联：

用户：分布式系统和机器学习之间有什么隐藏联系？
→ "你的'共识算法'和'反向传播'可能有关联：都通过迭代反馈达成全局一致性"

盲点检测：

用户：分析我对"神经网络"理解的盲点
→ "你了解 CNN、RNN、Transformer，但缺少：图神经网络、神经架构搜索、模型压缩..."

认知拓扑：

用户：展示我的知识图谱整体结构
→ 3 个孤岛、桥梁概念"一致性"、稀疏区域：系统安全和性能优化

灵感碰撞：

用户：碰撞"分布式系统"和"神经科学"
→ "大脑的神经可塑性类似于分布式系统的自适应拓扑。突触修剪 ≈ 节点退役。"

架构

MCP 客户端 (Cursor / CodeBuddy / Cline)
        │ stdio (JSON-RPC)
        │ + sampling/createMessage (Sampling 模式)
        ▼
┌──────────────────────────────────────┐
│  Cognitive Exoskeleton MCP Server   │
│                                      │
│  8 个 MCP 工具                       │
│         │                            │
│  知识图谱引擎 (SQLite + 图算法)       │
│         │                            │
│  LLM 双模式: Sampling / Direct       │
└──────────────────────────────────────┘

知识图谱数据模型

nodes (id, type, name, summary, domain, source_file,
       first_seen_at, last_seen_at, mention_count)
edges (id, source_id, target_id, relation, confidence, evidence, created_at)
notes_index (file_path, content_hash, node_ids, last_ingested_at)
evolution_log (id, node_id, snapshot_at, belief_summary, trigger_note, source_file)
topology_cache (snapshot_at, isolated_clusters, bridge_nodes, density_map, summary)
serendipity_log (id, node_a, node_b, hypothesis, user_feedback, created_at)

---

English

Cognitive Exoskeleton is not just a search tool. It builds a dynamic knowledge graph from your notes, then uses LLM reasoning to proactively discover blindspots, find hidden cross-domain connections, trace how your understanding evolves over time, and spark creative inspiration by colliding ideas from different fields.

• All data stays local (SQLite) — privacy-first
• Zero-config: uses your MCP client's LLM via Sampling protocol — or bring your own API (Hy3, OpenAI, Ollama, vLLM, etc.)
• Plug-and-play: compatible with Cursor, CodeBuddy, WorkBuddy, Cline, and other MCP clients

Features

8 MCP tools organized in four layers:

Layer	Tool	What it does
Foundation	ingest_note	Extract entities + relationships from notes into the knowledge graph
	query_mind	Answer questions using your knowledge graph (shallow/deep retrieval)
	recall_context	Surface forgotten notes related to what you're writing
Reasoning	discover_connections	Find hidden connections between knowledge from different domains
	detect_blindspots	Identify gaps, contradictions, and missing perspectives
	analyze_cognitive_topology	Generate a "cognitive portrait" — islands, bridges, dense/sparse regions
Temporal	trace_concept_evolution	Track how your understanding of a concept changes over time
Inspiration	spark_serendipity	Create creative sparks by colliding concepts from different domains

Quick Start

Prerequisites: Node.js >= 18

git clone https://github.com/Tencent-Hunyuan/Hy3.git
cd Hy3/rhinobird2026/cognitive-exoskeleton-mcp
npm install && npm run build

# Zero-config — automatically reuses your MCP client's LLM via Sampling
node dist/index.js

Zero-config mode: By default, the MCP Server delegates LLM calls to the client (Cursor, WorkBuddy, etc.) via MCP Sampling protocol. No separate API key needed.

For a standalone LLM API, switch to Direct mode:

export LLM_MODE=direct
export LLM_API_BASE="http://127.0.0.1:8000/v1"
export LLM_API_KEY="EMPTY"
export LLM_MODEL_NAME="hy3"
node dist/index.js

Environment Variables

Variable	Description	Default
LLM_MODE	LLM mode: sampling (delegate to client) or direct (API)	auto-detected*
LLM_API_BASE	(Direct) OpenAI-compatible API base URL	http://127.0.0.1:8000/v1
LLM_API_KEY	(Direct) API key for the LLM provider	EMPTY
LLM_MODEL_NAME	(Direct) Model name to use	gpt-4o-mini
COGNITIVE_DB_PATH	SQLite database file path	./cognitive.db

* Auto-detection: if LLM_API_BASE and LLM_API_KEY are both set (and key is not EMPTY), uses direct; otherwise uses sampling.

Multi-Model Support (Direct mode only)

Only needed when LLM_MODE=direct:

Provider	LLM_API_BASE	LLM_MODEL_NAME
Hy3 (local vLLM)	http://127.0.0.1:8000/v1	hy3
Hy3 (official API)	https://api.hunyuan.tencent.com/v1	hy3
OpenAI	https://api.openai.com/v1	gpt-4o
Ollama (local)	http://localhost:11434/v1	qwen2.5:32b
vLLM (any model)	http://127.0.0.1:8000/v1	<model-name>

MCP Client Setup

Cursor — .cursor/mcp.json (zero-config, reuses Cursor's model):

{
  "mcpServers": {
    "cognitive-exoskeleton": {
      "command": "node",
      "args": ["<project-path>/dist/index.js"],
      "env": {
        "LLM_MODE": "sampling"
      }
    }
  }
}

CodeBuddy / WorkBuddy — CLI command:

codebuddy mcp add cognitive-exoskeleton \
  --command "node" \
  --arg "<project-path>/dist/index.js" \
  --env LLM_MODE=sampling

Sampling mode: The MCP Server delegates LLM calls to the client via the MCP Sampling protocol. The client uses its own configured model (e.g. Cursor's Claude/GPT). No separate API key needed.

Usage Examples

Ingest a note:

User: Ingest this note: "Distributed systems follow the CAP theorem..."
→ Extracts CAP Theorem, Consistency, Availability, etc. + relationships

Graph Q&A:

User: What do I know about the CAP theorem?
→ Retrieves related entities, LLM reasons and returns structured answer

Writing recall:

User: I'm writing about database consistency models...
→ Recalls notes from 3 months ago about CAP theorem

Hidden connections:

User: Hidden connections between distributed systems and ML?
→ "Your 'consensus algorithms' and 'backpropagation' may be related:
   both achieve global consistency through iterative feedback"

Blindspot detection:

User: Blindspots in my understanding of neural networks?
→ "You know CNNs, RNNs, Transformers, but missing: GNNs, NAS, model compression..."

Cognitive topology:

User: Show me the overall structure of my knowledge graph
→ 3 islands, bridge concept "consistency", sparse: security, optimization

Serendipity spark:

User: Spark between distributed-systems and neuroscience
→ "Neural plasticity ≈ adaptive topology. Synaptic pruning ≈ node decommissioning."

Architecture

MCP Client (Cursor / CodeBuddy / Cline)
        │ stdio (JSON-RPC)
        ▼
┌──────────────────────────────────────┐
│  Cognitive Exoskeleton MCP Server   │
│                                      │
│  8 MCP Tools                         │
│         │                            │
│  Knowledge Graph Engine              │
│  (SQLite + graph algorithms)         │
│         │                            │
│  LLM Client                          │
│  (Model-agnostic, OpenAI-compatible) │
└──────────────────────────────────────┘

Knowledge Graph Schema

nodes (id, type, name, summary, domain, source_file,
       first_seen_at, last_seen_at, mention_count)
edges (id, source_id, target_id, relation, confidence, evidence, created_at)
notes_index (file_path, content_hash, node_ids, last_ingested_at)
evolution_log (id, node_id, snapshot_at, belief_summary, trigger_note, source_file)
topology_cache (snapshot_at, isolated_clusters, bridge_nodes, density_map, summary)
serendipity_log (id, node_a, node_b, hypothesis, user_feedback, created_at)

Development

npm install          # Install dependencies
npm run dev          # Watch mode (auto-rebuild)
npm run build        # Production build
node dist/index.js   # Start server

Tech Stack

Component	Choice	Notes
Language	TypeScript	Node.js >= 18
MCP SDK	@modelcontextprotocol/sdk	Official TypeScript SDK
Database	SQLite (sql.js)	Pure JS/WASM, zero native deps
LLM	openai SDK	OpenAI-compatible, any model
Markdown	gray-matter	Frontmatter parsing
Bundler	tsup	Single-file bundle

Demo Walkthrough

npm run build

# In your MCP client:
# 1. ingest_note → "examples/sample-notes/distributed-systems.md"
# 2. ingest_note → "examples/sample-notes/neural-networks.md"
# 3. query_mind → "What do I know about consensus?"
# 4. detect_blindspots → topic = "distributed systems"
# 5. analyze_cognitive_topology → (no arguments)
# 6. discover_connections → topic = "consensus"
# 7. spark_serendipity → domain_a = "distributed-systems", domain_b = "machine-learning"

---

License / 许可证

Apache-2.0

本项目为 2026 犀牛鸟开源人才培养活动参赛项目，基于腾讯混元 Hy3 模型构建。

This project was developed for the 2026 Rhinobird Open Source Talent Program, built on Tencent Hunyuan Hy3.

---

Copyright (c) 2026 hanjiang-215. All rights reserved.

本项目由 hanjiang-215 制作。