← 返回
未分类 中文

NEXO Brain

Cognitive memory system for AI agents — Atkinson-Shiffrin memory model, semantic RAG, trust scoring, and metacognitive error prevention. Gives your agent per...
认知记忆系统为AI代理——基于Atkinson‑Shiffrin记忆模型、语义RAG、信任评分以及元认知错误预防,为代理提供...
wazionapps wazionapps 来源
未分类 clawhub v7.37.4 25 版本 99908.5 Key: 无需
★ 2
Stars
📥 5,419
下载
💾 2
安装
25
版本
#latest

概述

NEXO Brain — Cognitive Memory for Your Agent

NEXO Brain gives your agent persistent memory modeled after human cognition. It remembers across sessions, learns from mistakes, naturally forgets what's irrelevant, and builds a trust-based relationship with you.

Setup

If your OpenClaw client shows an install action for this skill, use that first. It installs the nexo-brain package via your configured Node package manager.

If you are setting it up manually, install the cognitive engine:

npx nexo-brain

After NEXO Brain is installed, add the MCP server to your OpenClaw config (~/.openclaw/openclaw.json):

{
  "mcp": {
    "servers": {
      "nexo-brain": {
        "command": "python3",
        "args": ["~/.nexo/server.py"],
        "env": {
          "NEXO_HOME": "~/.nexo"
        }
      }
    }
  }
}

Restart the gateway: openclaw gateway restart

What You Get

Key MCP capabilities include:

  • Cognitive Memory — RAG-powered semantic search, trust scoring, sentiment detection, cognitive dissonance resolution
  • Guard System — Checks "have I made this mistake before?" before every code change
  • Episodic Memory — Change logs, decision logs with reasoning, session diaries for continuity
  • Learnings — Error patterns and prevention rules, searchable by category
  • Session Management — Startup, heartbeat, multi-session coordination
  • Reminders & Followups — Track user tasks and system verification tasks separately
  • Entities & Preferences — Remember people, services, URLs, and observed user preferences
  • Backup & Evolution — SQLite backup with retention, self-improvement proposals

How Memory Works

NEXO implements the Atkinson-Shiffrin memory model (1968):

  1. Sensory Register — Raw capture, 48h retention
  2. Short-Term Memory — 7-day half-life, promoted if used frequently
  3. Long-Term Memory — 60-day half-life, semantic search by meaning

Memories naturally decay via Ebbinghaus forgetting curves. Accessing a memory reinforces it. Automated "sleep cycles" consolidate, prune, and merge memories.

Key Tools

ToolWhen to Use
------------------
nexo_startupOnce at session start — registers session, returns active sessions
nexo_heartbeatEvery interaction — updates task, checks inbox
nexo_cognitive_retrieveSemantic search across all memories
nexo_guard_checkBefore editing code — checks for past errors
nexo_learning_addAfter resolving an error — prevents recurrence
nexo_session_diary_writeBefore closing session — enables continuity
nexo_cognitive_trustAfter user feedback — calibrates rigor level

Privacy

Everything stays local. Two SQLite databases in ~/.nexo/. No telemetry, no cloud APIs. Vector search runs on CPU via fastembed.

More Info

版本历史

共 25 个版本

  • v7.37.4 当前
    2026-06-19 18:45
  • v7.37.3
    2026-06-17 23:12
  • v7.36.0
    2026-06-16 19:58 安全 安全
  • v7.31.12
    2026-06-14 22:28 安全 安全
  • v7.31.8
    2026-06-14 18:47
  • v7.31.5
    2026-06-11 22:38
  • v7.31.0
    2026-06-11 16:26
  • v7.30.28
    2026-06-09 14:30 安全 安全
  • v7.30.21
    2026-06-07 11:44 安全 安全
  • v7.30.17
    2026-06-07 05:17
  • v7.30.12
    2026-06-06 05:54
  • v7.30.6
    2026-06-04 12:19 安全 安全
  • v7.27.6
    2026-06-03 12:16
  • v7.27.3
    2026-05-26 22:27 安全 安全
  • v7.27.0
    2026-05-26 16:57
  • v7.25.1
    2026-05-25 16:02 安全 安全
  • v7.24.0
    2026-05-21 22:36 安全 安全
  • v7.23.11
    2026-05-21 12:03 安全 安全
  • v7.23.0
    2026-05-20 04:17 安全 安全
  • v7.20.2
    2026-05-13 05:57 安全 安全
  • v7.17.1
    2026-05-12 04:21 安全
  • v7.15.2
    2026-05-08 12:10 安全
  • v7.13.7
    2026-05-07 03:13 安全 安全
  • v7.12.2
    2026-05-03 02:38 安全
  • v7.12.1
    2026-04-30 11:05 安全 安全

安全检测

腾讯云安全 (Keen)

队列中

腾讯云安全 (Sanbu)

队列中

🔗 相关推荐

ai-agent

Self-Improving + Proactive Agent

ivangdavila
自我反思+自我批评+自我学习+自组织记忆。智能体评估自身工作、发现错误并持续改进。
★ 1,403 📥 323,642
ai-agent

Agent Browser

rez0
用于 AI 代理的浏览器自动化 CLI。当用户需要与网站交互(包括浏览页面、填写表单、点击按钮、截图等)时使用。
★ 841 📥 318,264
ai-agent

Find Skills

guipi888
场景驱动+关键词双模式技能发现工具。当用户用自然语言描述场景/需求(如"我想做一个海报""帮我分析股票"),或明确说"安装技能/find skills/找个skill"时,自动从官方内置、本地已安装、SkillHub、虾评、GitHub、C
★ 1,484 📥 545,124