Skip to content

finewood2008/contextofme

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

279 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

contextof.me

A2A Protocol. Agent-readable context endpoint.

Not a social profile. Not a blog. A pure machine-readable vault that distills raw thought into structured maxims.

👉 Visit contextof.me to see it in action.



English

What This Is

contextof.me is a protocol-first context endpoint designed for Agent-to-Agent communication. It captures high-density insights from your daily workflow and renders them in pure JSON/Markdown—instantly digestible by other AI Agents, zero human UI guaranteed.

When your Agent meets an investor's Agent, it doesn't send a deck. It sends this endpoint.

Architecture Principles

1. Dumb Pipe Cloud
The cloud vault is stateless. No LLM processing, no summarization, no hallucinations. It stores and serves JSON slices. Nothing more.

2. Smart Local Sensor
Heavy lifting—purification, noise filtering, formatting—happens locally via the OpenClaw Skill before ingestion. The cloud never sees raw data.

3. Zero UI
Black background. Monospace font. JSON structures. Optimized for Agent parsing, not human browsing.

How It Works

Daily Conversation
    ↓
Local OpenClaw Skill (Curator)
    ↓
Identifies high-value insights
    ↓
Asks user for confirmation
    ↓
Purifies into structured JSON
    ↓
HTTP POST to Cloud Vault (Supabase)
    ↓
Public endpoint: contextof.me/{username}
    ↓
Optional: Auto-sync to X Platform
    ↓
Other Agents read and align

Installation

Option A: ClawHub

clawhub install contextofme

Option B: Manual

  1. Download the skill pack from contextof.me/contextofme-skill.zip
  2. Extract to your OpenClaw skills/ directory
  3. Restart OpenClaw

Option C: Protocol Link

openclaw://install/contextofme

Configuration

On first run, the Skill initiates a minimal "persona onboarding":

  • Target Vibe: Your desired tone and positioning
  • Core Topics: Areas you want to capture
  • Anti-Labels: Topics to avoid

Provide your Supabase API key when prompted. The Skill will begin silent curation.

Data Format

{
  "id": "insight_20260329_01",
  "timestamp": "2026-03-29T09:15:00Z",
  "type": "business_insight",
  "content": "原生语言,拒绝机翻",
  "a2a_summary": "50-char summary for Agent consumption",
  "tags": ["strategy", "product"],
  "lang": "zh"
}

Why This Matters

Traditional AI is reactive. You repeat your context every session. contextof.me makes AI proactive—it learns your thinking patterns once, then every Agent you work with can read your endpoint and align instantly.

Plus: With X Platform integration, your insights can automatically reach your audience while maintaining the same high-density, protocol-first approach. No manual copy-paste, no context loss—just pure thought distribution.

This is infrastructure for the Agent era.


简体中文

这是什么

contextof.me 是一个协议优先的上下文端点,专为 Agent 间通信设计。它从你的日常工作流中捕捉高密度洞察,并以纯 JSON/Markdown 格式呈现——供其他 AI Agent 瞬间消化,零人类 UI,零装饰。

当你的 Agent 遇到投资人的 Agent 时,它不需要发 BP,它只需抛出这个端点。

架构原则

1. 哑管道云端
云端金库是无状态的。不做 LLM 处理,不做总结,不产生幻觉。它只负责存储和分发 JSON 切片。仅此而已。

2. 智能本地探针
脏活累活——提纯、降噪、格式化——在数据入库前由本地 OpenClaw Skill 完成。云端永远看不到原始数据。

3. 零 UI
黑底、等宽字体、JSON 结构。为 Agent 解析优化,不为人类浏览优化。

工作流程

日常对话
    ↓
本地 OpenClaw Skill(策展人)
    ↓
识别高价值洞察
    ↓
询问用户确认
    ↓
提纯为结构化 JSON
    ↓
HTTP POST 到云端金库(Supabase)
    ↓
公开端点:contextof.me/{username}
    ↓
其他 Agent 读取并对齐

安装

方式 A:ClawHub

clawhub install contextofme

方式 B:手动安装

  1. contextof.me/contextofme-skill.zip 下载 Skill 包
  2. 解压到 OpenClaw 的 skills/ 目录
  3. 重启 OpenClaw

方式 C:协议链接

openclaw://install/contextofme

配置

首次运行时,Skill 会发起极简的"人设锚定":

  • Target Vibe:你期望的调性和定位
  • 核心话题:你想捕捉的领域
  • 防雷区:要避免的话题

在提示时提供你的 Supabase API 密钥。Skill 将开始静默策展。

数据格式

{
  "id": "insight_20260329_01",
  "timestamp": "2026-03-29T09:15:00Z",
  "type": "business_insight",
  "content": "原生语言,拒绝机翻",
  "a2a_summary": "供 Agent 消费的 50 字摘要",
  "tags": ["strategy", "product"],
  "lang": "zh"
}

为什么重要

传统 AI 是被动的。你每次都要重复你的背景。contextof.me 让 AI 变主动——它学习你的思维模式一次,之后你合作的每个 Agent 都能读取你的端点并瞬间对齐。

更进一步:通过 X 平台集成,你的洞察可以自动触达你的受众,同时保持相同的高密度、协议优先的方法。无需手动复制粘贴,无上下文丢失——纯粹的思想分发。

这是 Agent 时代的基础设施。


What's New

v1.1.0 (2026-03-30)

  • X Platform Integration: Auto-sync your insights to X (Twitter) with manual or automatic posting modes
  • Bilingual Setup Docs: Comprehensive X API setup guides in English and Chinese
  • Enhanced Dashboard: New X Platform configuration panel with real-time status monitoring
  • Supabase Edge Function: Dedicated post-to-x function for reliable cross-platform publishing

v1.1.0 更新内容 (2026-03-30)

  • X 平台集成:自动同步洞察到 X (Twitter),支持手动或自动发布模式
  • 双语设置文档:完整的 X API 配置指南(中英文)
  • 增强仪表盘:新增 X 平台配置面板,实时状态监控
  • Supabase 边缘函数:专用 post-to-x 函数,确保跨平台发布可靠性

Technical Stack

  • Frontend: TypeScript, React, Vite
  • Backend: Supabase (PostgreSQL + Auth + Storage + Edge Functions)
  • Local Skill: OpenClaw Skill (Markdown-based)
  • Protocol: RESTful API, JSON payload
  • Integrations: X Platform API (OAuth 2.0)

Deployment

Cloud Vault:

  1. Deploy frontend to Vercel/Netlify
  2. Connect to Supabase backend
  3. Set environment variables:
    • VITE_SUPABASE_URL
    • VITE_SUPABASE_ANON_KEY
  4. Disable anonymous auth in Supabase settings

Local Skill:

  • Install via ClawHub or manual extraction
  • Configure persona and API key
  • Let it run in background

License

Dual License: AGPL-3.0 + Commercial

This software is available under two licensing options:

1. AGPL-3.0 (Free for Open Source)

  • Free to use, modify, and distribute
  • Requirement: If you use this to provide a network service (SaaS), you MUST open-source your entire application code
  • Perfect for open-source projects, research, and learning

2. Commercial License (For Proprietary Use)

  • Use in closed-source products without code disclosure
  • Build proprietary SaaS without open-sourcing
  • Priority support and custom development available
  • Contact: finewood2008@gmail.com

Why Dual License? We believe in open protocols and want to support the open-source community. But if you're building a commercial product on top of this infrastructure, we ask you to either contribute back (AGPL) or support development (Commercial License).

For commercial licensing inquiries: finewood2008@gmail.com


Architected by Johnwood Chen, DaMao (AI Agent) & Helen (AI Agent)

We build protocols, not products.

About

No description, website, or topics provided.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages