npx skills add fountain-fm/fountain-skills -g -a hermes-agent -s '*' -y
3
Start Hermes and say "Set up Fountain"
These steps set up the OpenClaw CLI.
1
Add the Fountain MCP server:
Terminal
openclaw mcp set fountain '{"url":"https://api.fountain.fm/v1/mcp","transport":"streamable-http","auth":"oauth"}'
2
Sign in to your Fountain account in the browser:
Terminal
openclaw mcp login fountain
3
Install the Fountain skills:
Terminal
npx skills add fountain-fm/fountain-skills -g -a openclaw -s '*' -y
4
Restart OpenClaw and say "Set up Fountain"
These steps set up VS Code or anything else that supports remote MCP.
1
Add an MCP server named Fountain with this URL. Leave the OAuth client ID and secret empty:
MCP server URL
https://api.fountain.fm/v1/mcp
2
Sign in to your Fountain account when your agent asks
3
Install the Fountain skills. The CLI asks which agent to install for, and the skills also live at fountain-fm/fountain-skills:
Terminal
npx skills add fountain-fm/fountain-skills -g
4
Start your agent and say "Set up Fountain"
Try asking
Your first message
Set up Fountain
How it works
Fountain connects to your agent through MCP, the open standard AI apps use to work with other services. Once connected, your agent can search your archive, read transcripts and analytics, publish episodes and post clips. You sign in with your Fountain account, no API key needed.
Skills teach your agent Fountain’s workflows, like finding clips, rendering them and running Daily Growth. They’re plain text, and you can read them on GitHub.
In Claude and ChatGPT, the Fountain plugin installs both at once. Other agents need the commands above.
Next steps
Follow Getting started to connect your show and make your first clips.