every ai you use

innernet works wherever you talk to an ai. one memory, kept current by netti, and every app below reads and writes the same one. tell one of them something on monday and the next one knows it on tuesday.

each app takes innernet in one of four ways: a plugin you add with one press, a link you paste as a connector, a block for a config file, or a command in a terminal. every one of them signs in through your browser. there is never a key or a password to paste into a chat.

the shortest way, in any ai#

paste these three lines into the ai you want to connect. it reads the guide written for it and walks you through its own settings, one step at a time.

connect innernet, my memory, to this ai.
read https://innernet.live/connect/ai and follow it: set it up yourself if you can, or walk me through it one step at a time.
the link is https://innernet.live/api/mcp

on a computer with a terminal, npx -y innernet connect signs in once and wires every ai tool it finds there.

at a glance#

innernet is also in the official MCP registry as live.innernet/innernet, so any app that reads the registry can find it by name.

Claude#

one url, pasted as a connector.

1. in Claude, open Customize → Connectors and press Add
2. name it innernet, paste this link, press Continue
3. press Connect, then allow access in the tab that opens

the link:

https://innernet.live/api/mcp

it goes in: Customize → Connectors → Add custom connector → Remote MCP server URL.

so it stops asking every time: open Customize → Connectors → innernet, then set both tool groups to Always allow.

the standing instruction goes in Settings → General → Instructions for Claude.

for your ai, the same steps written for a model: connect innernet to Claude

ChatGPT#

one press · the innernet plugin.

add the innernet plugin

1. press "add the innernet plugin"
2. on the innernet page in ChatGPT, press Connect
3. sign in to innernet and allow access

so it stops asking every time: when ChatGPT asks before a write, choose "remember for this conversation".

the standing instruction goes in Settings → Personalization → Custom instructions.

for your ai, the same steps written for a model: connect innernet to ChatGPT

Grok Bot#

ask your Bot, then Connect.

1. in Grok Bot, open the Bot that should remember and say: add the custom MCP server innernet at this link
2. press Connect on the card it shows, then sign in to innernet and allow it
3. start a new task. innernet is under Connect apps → Installed

the link:

https://innernet.live/api/mcp

it goes in: a message to your Bot (it answers with a connect card).

the standing instruction goes in the Bot's job, in its details.

for your ai, the same steps written for a model: connect innernet to Grok Bot

Grok#

one url, as a custom connector.

1. open grok.com/connectors and press New Connector
2. choose Custom, name it innernet and paste this link
3. sign in when your browser opens

the link:

https://innernet.live/api/mcp

it goes in: grok.com/connectors → New Connector → Custom.

the standing instruction goes in Grok's custom instructions.

for your ai, the same steps written for a model: connect innernet to Grok

Perplexity#

one url, as a custom connector.

1. on the web, open Plugins in the sidebar and press Add custom
2. choose remote, name it innernet, paste this link and pick OAuth
3. sign in when your browser opens

the link:

https://innernet.live/api/mcp

it goes in: Plugins → Add custom → remote server url.

the standing instruction goes in your Perplexity profile instructions.

for your ai, the same steps written for a model: connect innernet to Perplexity

Le Chat#

mistral · one url, as a connector.

1. open Connectors and press + Add Connector
2. choose Custom MCP Connector, name it innernet and paste this link
3. sign in when your browser opens

the link:

https://innernet.live/api/mcp

it goes in: Connectors → + Add Connector → Custom MCP Connector.

the standing instruction goes in Le Chat's custom instructions.

for your ai, the same steps written for a model: connect innernet to Le Chat

Notion#

custom agents · business plans.

1. open a custom agent's settings → Tools & Access
2. press Add connection → Custom MCP server and paste this link
3. sign in when your browser opens

the link:

https://innernet.live/api/mcp

it goes in: custom agent → Tools & Access → Custom MCP server.

the standing instruction goes in Notion AI's custom instructions.

for your ai, the same steps written for a model: connect innernet to Notion

the Gemini app#

one link · as a custom app.

1. on the web, open gemini.google.com → Settings → Connected Apps
2. press Add a custom app, name it innernet and paste the link below
3. sign in when your browser opens, then type @innernet in a chat

https://innernet.live/api/mcp

it goes in: gemini.google.com → Settings → Connected Apps → Add a custom app.

Claude Code#

one command, cleanest.

1. open a terminal
2. run the command below. it signs you in and wires every tool it finds

in a terminal:

npx innernet connect

it goes in: any terminal on the machine.

so it stops asking every time: type /permissions in Claude Code, then choose Allow and add mcp__innernet__*.

the standing instruction goes in your project's CLAUDE.md.

for your ai, the same steps written for a model: connect innernet to Claude Code

Codex#

one press · the innernet plugin.

add the innernet plugin

1. press "add the innernet plugin"
2. press Connect and sign in to innernet
3. start a new thread in Codex; innernet is in its plugins

the standing instruction goes in your AGENTS.md.

for your ai, the same steps written for a model: connect innernet to Codex

Gemini CLI#

google's terminal agent.

1. open ~/.gemini/settings.json
2. add the block below under mcpServers
3. restart gemini, run /mcp auth innernet, and sign in

the block, beside any servers already there:

{
  "mcpServers": {
    "innernet": {
      "httpUrl": "https://innernet.live/api/mcp"
    }
  }
}

it goes in: ~/.gemini/settings.json.

the standing instruction goes in your GEMINI.md.

for your ai, the same steps written for a model: connect innernet to Gemini CLI

Hermes Agent#

nous research · one block, one login.

1. open ~/.hermes/config.yaml
2. add the block below under mcp_servers and save
3. run hermes mcp login innernet and sign in when your browser opens

the block, beside any servers already there:

mcp_servers:
  innernet:
    url: "https://innernet.live/api/mcp"
    auth: oauth

it goes in: ~/.hermes/config.yaml.

the standing instruction goes in your AGENTS.md.

for your ai, the same steps written for a model: connect innernet to Hermes Agent

OpenClaw#

the open agent · two commands.

1. open a terminal on the machine OpenClaw runs on
2. run the first command below, then the second, and sign in when your browser opens

in a terminal, one after the other:

openclaw mcp set innernet '{"url":"https://innernet.live/api/mcp","transport":"streamable-http","auth":"oauth"}'
openclaw mcp login innernet

it goes in: a terminal where openclaw runs.

the standing instruction goes in your agent's AGENTS.md.

for your ai, the same steps written for a model: connect innernet to OpenClaw

Cursor#

one press, then Install.

open in Cursor on the machine it runs on.

1. press "open in Cursor"
2. press Install
3. sign in when your browser opens

by hand, the block, beside any servers already there:

{
  "mcpServers": {
    "innernet": {
      "url": "https://innernet.live/api/mcp"
    }
  }
}

it goes in: ~/.cursor/mcp.json (or .cursor/mcp.json in one project).

so it stops asking every time: in Cursor, open Settings → Agents, then turn on auto-run for MCP tools.

the standing instruction goes in your project rules (.cursor/rules).

for your ai, the same steps written for a model: connect innernet to Cursor

Antigravity#

one block in mcp_config.json.

1. in the agent panel, open … → MCP Servers → Manage MCP Servers
2. press View raw config, add the block below under mcpServers, and save
3. sign in when your browser opens

the block, beside any servers already there:

{
  "mcpServers": {
    "innernet": {
      "serverUrl": "https://innernet.live/api/mcp"
    }
  }
}

it goes in: ~/.gemini/config/mcp_config.json.

the standing instruction goes in your agent instructions.

for your ai, the same steps written for a model: connect innernet to Antigravity

Windsurf#

cascade reads one config file.

1. in Cascade, open the … menu → MCPs → Open MCP config file
2. add the block below under mcpServers and save
3. sign in when your browser opens

the block, beside any servers already there:

{
  "mcpServers": {
    "innernet": {
      "serverUrl": "https://innernet.live/api/mcp"
    }
  }
}

it goes in: mcp_config.json, the file Cascade opens.

the standing instruction goes in your Windsurf rules.

for your ai, the same steps written for a model: connect innernet to Windsurf

VS Code and GitHub Copilot#

github copilot, one press.

open in VS Code on the machine it runs on.

1. press "open in VS Code"
2. press Install
3. sign in when your browser opens

by hand, the link:

https://innernet.live/api/mcp

it goes in: Command Palette → MCP: Add Server → HTTP.

so it stops asking every time: when Copilot asks before a tool call, choose "always allow".

the standing instruction goes in your .github/copilot-instructions.md.

for your ai, the same steps written for a model: connect innernet to VS Code and GitHub Copilot

Cline#

the VS Code agent.

1. open MCP Servers → Remote Servers
2. name it innernet and paste this link
3. set the transport to Streamable HTTP and press Add Server
4. sign in when your browser opens

the link:

https://innernet.live/api/mcp

it goes in: MCP Servers → Remote Servers → Server URL.

the standing instruction goes in your Cline custom instructions.

for your ai, the same steps written for a model: connect innernet to Cline

Continue#

one block in config.yaml.

1. open ~/.continue/config.yaml
2. add the block below. if mcpServers is already there, add the item under it
3. reload Continue and sign in when your browser opens

the block, beside any servers already there:

mcpServers:
  - name: innernet
    type: streamable-http
    url: https://innernet.live/api/mcp

it goes in: ~/.continue/config.yaml.

the standing instruction goes in your Continue system message.

for your ai, the same steps written for a model: connect innernet to Continue

Zed#

the fast editor.

1. open Settings → AI → MCP Servers → Add Server → Add Remote Server
2. name it innernet and paste this link
3. sign in when your browser opens

the link:

https://innernet.live/api/mcp

it goes in: Add Remote Server → url (settings.json context_servers).

the standing instruction goes in your Zed assistant rules.

for your ai, the same steps written for a model: connect innernet to Zed

any other MCP client#

one url, any client that speaks MCP.

1. open your tool's MCP settings
2. add a remote server and paste this link
3. sign in when your browser opens

the link:

https://innernet.live/api/mcp

the standing instruction goes in your tool's standing instructions.

not yet#

Meta AI, the Microsoft Copilot consumer app, DeepSeek and Instinct cannot hold a memory like this yet. connect an app that can, and what you keep there is waiting for the day they open up.

an app missing from this page that speaks MCP takes the same link: look for remote server, custom connector or MCP in its settings.