Skip to main content
NanoClaw is an open-source assistant that runs each agent in its own Docker container on your machine, built on the Claude Agent SDK. You talk to it over Slack, Telegram, Discord, WhatsApp, iMessage and other chat apps. Add PostQueen’s MCP server to an agent group and you can ask it, from any of those apps, to write your posts for PostQueen to schedule, and check how they did.
A sample NanoClaw chat on a phone with PostQueen connected: a request for posts, the PostQueen tools it used, and the reply

Before you start

  • Plans: NanoClaw is free and open source (MIT). You run it yourself with an Anthropic API key or a Claude subscription.
  • Platforms: your own macOS or Linux machine, or Windows through WSL2, with Docker. The installer adds Node 22, pnpm and Docker if they are missing.
  • NanoClaw, installed, with at least one agent group. Follow NanoClaw’s install guide.
  • A PostQueen API key. A workspace admin copies it from Connections.
Where the API key is: in the app, open Connections > API key. Only a workspace Admin or Super Admin can reveal or rotate it; other members see it masked. A workspace has one key, and each agent card on Connections shows your MCP address with the key already filled in.

Connect NanoClaw

NanoClaw keeps keys in its OneCLI vault and adds them to the agent’s requests, so the agent itself never sees your key. Run these on the machine that runs NanoClaw.
1

Save the key in the vault

2

Find your agent group

3

Check that the group's agent can use the key

Each agent group has a OneCLI agent whose identifier is the group’s id; it appears once the group has run. On the OneCLI version NanoClaw installs, agents use secret mode all and pick up the PostQueen secret by its host, so there is nothing to grant. If the agent shows secretMode selective, add the secret (its id is in onecli secrets list) as NanoClaw’s Credentials page shows. On OneCLI with agent grants (server 1.44 or later, CLI 2.9 or later):
4

Add PostQueen and restart the group

Use the group’s id in place of YOUR_GROUP_ID. The address has no key in it:

Check that it works

Message the agent in one of your chat apps: “List my PostQueen channels.” It answers with the channels you connected in PostQueen. If the tools do not load, see Troubleshooting.

Good to know

  • Last resort if the tools still do not load: add the server with the key as a header instead, then restart the group. This goes against NanoClaw’s advice: --headers is meant for non-secret headers, and real credentials belong in the vault.
    The key then sits in the group’s config, where the agent can read it.
  • Do not use the address with the key in it. The key would sit in the group’s config. Keep it in the OneCLI vault.
  • Use the MCP server, not the PostQueen CLI or skill. NanoClaw keeps keys out of its containers: it rejects credential values in a container’s environment. The MCP server gets the key from the vault.
  • The agent can add the server itself with its add_mcp_server tool, which needs admin approval in NanoClaw. The key still has to go into the vault first.
  • Ask for drafts when you want to check first. A draft waits on your PostQueen calendar, and nothing publishes until you schedule it.
  • What it cannot do over MCP: change a post’s text or time. Do that on the calendar in the app. The MCP tools lists what it can do.

Other ways to connect

Try this first

The drafts appear on your PostQueen calendar, ready for you to check and schedule. Prompts to try has more.

Run it on a schedule

Ask your agent in chat for a recurring task, such as the one below. It creates the task with ncl tasks create, and ncl tasks list shows every series.
Source: NanoClaw, scheduled tasks.

Troubleshooting

More fixes are on Troubleshooting agents. To pick a chat app, see Chat channels.

Is it the same as OpenClaw?

No. The names are close, but they are different apps with different setups:

Next steps

Every agent

Which agents can post to PostQueen, and how each one connects.

The MCP tools

What each of the 25 tools does, and what an agent cannot do.

Prompts and tool calls

The order an agent calls the tools in, with prompts that work.

Troubleshooting agents

What to do when a connection or a command fails.
Last modified on October 27, 2026