trip2g
Self-hosted knowledge mesh with built-in MCP server enabling AI agents to search and retrieve notes across federated hubs.
README
trip2g
Open-source MCP knowledge mesh. Self-host your knowledge bases, expose them to AI agents via MCP, and federate with peers — no SaaS in the middle.
Unlike Obsidian Publish or Quartz, trip2g is a live server: subscriptions, webhook agents, federation between hubs, and an MCP endpoint out of the box.
Try the public hub → · Self-host · Docs

Second brain network for your agents
A second brain isn't one person's vault — it's the shared knowledge of people who think about the same things.
Each person runs a hub with their own notes. Agents reach into it directly via MCP. Hubs peer with each other, so one question fans out across the network — your bases plus the ones your trusted peers share with you.
human A human B human C
│ │ │
▼ ▼ ▼
agent A agent B agent C
│ │ │
▼ ▼ ▼
┌────────┐ MCP ┌────────┐ MCP ┌────────┐
│ hub A │ ◄───► │ hub B │ ◄───► │ hub C │
└────────┘ └────────┘ └────────┘
▲ ▲ ▲
└─── humans browse · agents query ──┘
Same hub serves the human (a website with subscriptions, RSS, Telegram) and the agent (MCP). Data lives where you put it. trip2g is a protocol, not a vendor.
Try it now
Add the public knowledge hub to your MCP client:
{
"mcpServers": {
"trip2g": {
"url": "https://trip2g.com/_system/mcp"
}
}
}
Ask your agent anything. It searches all connected bases and returns answers with sources.
Want your own? Free cloud instance — no terminal needed.
Core capabilities
Markdown → website
- Wikilinks (
[[note]]), backlinks, outlinks — global resolution, just like Obsidian - Composable page layouts via frontmatter: sidebars, magazine grid, TOC, custom note embeds
- Hybrid search: bleve full-text + OpenAI embeddings, merged via reciprocal rank fusion
- Custom domains, multi-language with
hreflang, RSS output, sitemap
MCP server — built into every hub:
| Tool | Purpose |
|---|---|
search |
Hybrid full-text + semantic search |
note_html |
Read a note (or a section) by id, path, or match |
similar |
Notes similar to a given note |
federated_search / federated_similar / federated_note_html |
Same, fanned out to peer hubs |
instructions |
Author-defined prompt for the agent |
Access is scoped to the caller's subscription. Custom tools can be defined in note frontmatter (mcp_method:).
Webhook agents
- Change webhooks — POST to an external agent on note create/update/remove. Agent writes notes back via API. Glob filtering, HMAC auth, depth tracking prevents recursion.
- Cron webhooks — run an agent on a schedule (
0 9 * * *). Sync or async. Optional instruction context.
Monetization
- Subgraph paywalls — group notes into paid products, free notes stay public
- Crypto payments (NowPayments), Patreon and Boosty integration
Obsidian plugin — one-click sync, hash-based diff, only changed files are uploaded.
Federation
graph LR
O[Obsidian vault] --> H[your hub]
T[Telegram] --> H
H <-->|MCP federation| H2[peer hub]
H2 --> B[their bases]
H -->|/_system/mcp| A[agent]
H2 -->|/_system/mcp| A
Peer hubs with trusted people or orgs. Each hub controls access per base. One agent question reaches the union of all connected knowledge.
| Topology | Setup | Result |
|---|---|---|
| Solo | One hub, many bases | All your notes, books, courses — one query |
| Friends | Each person runs a hub, hubs peer | Union of everyone's knowledge |
| Company | Central hub + per-employee hubs | Tribal knowledge and docs — queryable |
| B2B | Two star topologies, one bridge | Shared knowledge without merging systems |
Sources
| Source | Status |
|---|---|
| Obsidian | ready — vault stays local, two-way sync |
| Telegram | ready — channel publish + history mirror |
| RSS output | ready — every base exposes feeds |
| Notion | planned |
| Google Drive | planned |
| Linear, Slack archive, RSS import | planned |
Template system
Two paths — pick one per knowledge base.
A. Default template (no code, frontmatter-only). Built-in trip2g layout. Compose pages from widgets and content blocks via frontmatter:
---
header: "[[Navigation]]"
left_sidebar: [TOC, inlinks]
content: [selfcontent, magazine]
magazine_include_files: "blog/**/*.md"
footer: "[[Footer]]"
---
Rendered through quicktemplate. Notes can also render as HTML, JSON, RSS via content_type frontmatter.
B. Custom Jet templates (full control). Drop your own .html files into a layouts folder and switch via layout: path/to/template. Templates have access to the markdown AST — iterate sections, render specific parts, customize layout down to HTML. Built on the Jet template engine.
Self-host
docker compose up
Full guide → · MIT · SQLite, no external dependencies required.
Tech stack
| Backend | Go, FastHTTP, gqlgen (GraphQL) |
| Database | SQLite + Litestream for streaming backup |
| Search | bleve (full-text) + OpenAI embeddings (semantic) |
| Markdown | Goldmark — wikilinks, frontmatter |
| Templates | quicktemplate (default) + Jet (custom) |
| Frontend | $mol, TypeScript, Tiptap |
| Assets | S3-compatible (MinIO for dev) |
License
MIT
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。