Br8n MCP server
{
"mcpServers": {
"br8n": {
"command": "npx",
"args": [
"-y",
"@br8n/mcp@latest"
]
}
}
}{
"mcpServers": {
"br8n": {
"command": "npx",
"args": [
"-y",
"@br8n/mcp@latest"
]
}
}
}{
"mcpServers": {
"br8n": {
"command": "npx",
"args": [
"-y",
"@br8n/mcp@latest"
]
}
}
}{
"servers": {
"br8n": {
"command": "npx",
"args": [
"-y",
"@br8n/mcp@latest"
]
}
}
}{
"mcpServers": {
"br8n": {
"command": "npx",
"args": [
"-y",
"@br8n/mcp@latest"
]
}
}
}file .mcp.json (in your project root)file claude_desktop_config.jsonfile ~/.cursor/mcp.jsonfile .vscode/mcp.jsonfile ~/.codeium/windsurf/mcp_config.jsoninstall npx -y @br8n/mcpversion 0.1.2
What it does
br8n is an MCP server that exposes a markdown-based knowledge base stored as plain files you own. It lets any AI model read, search, and cite from your team's documented processes, decisions, and institutional knowledge without vendor lock-in.
- Query your team's documented processes and have the model cite the exact file and line number as its source.
- Migrate your AI working memory between different models or tools by copying the same markdown files to a new machine.
- Store decision rationale alongside decisions so models can identify conflicts between current requests and past documented choices.
Tools this server exposes
| Tool | Does |
|---|---|
| brain_list | Lists all files in the brain directory structure. |
| brain_read | Reads the full contents of a specified markdown file from the brain. |
| brain_search | Searches the brain files for text and returns matching results with file and line number references. |
Before you enable it
An MCP server runs with whatever access you give it. Start with a development account rather than production, and check which tools the server exposes before you let an agent call them unattended.