Unified Connection Management

puppyone conn, or puppyone connection, is PuppyOne's unified connection management command. Every external integration, including data-source sync, agents, MCP endpoints, sandboxes, and local folders, is managed through the same command family.

Command overview

Command	Description
`conn add <type> [source]`	Create a connection
`conn ls`	List all connections
`conn info <id>`	Show connection details
`conn pause <id>`	Pause a connection
`conn resume <id>`	Resume a connection
`conn rm <id>`	Delete a connection
`conn key <id>`	Show the access key
`conn refresh <id>`	Trigger sync manually
`conn trigger <id> <mode>`	Set trigger mode
`conn update <id>`	Update configuration
`conn logs <id>`	Show execution logs
`conn run <id>`	Run the connection once
`conn schema <provider>`	Show configuration fields

Create connections

conn add is the main entry point. The first argument selects the connection type, and the remaining arguments and flags depend on that type.

Connection types

Type	Description	Example
`notion`	Notion data source	`conn add notion <url>`
`github`	GitHub repository	`conn add github <url>`
`gmail`	Gmail inbox	`conn add gmail`
`url`	Web page scraping	`conn add url <url>`
`agent`	AI agent	`conn add agent "name"`
`mcp`	MCP endpoint	`conn add mcp "name"`
`sandbox`	Code sandbox	`conn add sandbox "name"`
`folder`	Local folder sync	`conn add folder <path>`

Common options

Option	Description
`--name <name>`	Connection name
`--folder <path>`	Cloud directory to store data
`--mode <mode>`	Sync mode
`--model <model>`	Model used by an agent
`--system-prompt <text>`	System prompt for an agent
`--type <type>`	Sandbox type, `e2b` or `docker`
`--api-key <key>`	API key
`--config <json>`	Advanced JSON configuration

Data source connections

Pull data from external platforms into your Content Node tree:

# Notion: sync one page
puppyone conn add notion https://notion.so/my-page-id --folder /docs
 
# GitHub: sync one repository
puppyone conn add github https://github.com/org/repo --folder /code
 
# Gmail: sync email
puppyone conn add gmail
 
# Web page scraping
puppyone conn add url https://example.com/article --folder /imports

Agent connections

Create an AI agent and bind its model and tools:

# Create a GPT-4o agent
puppyone conn add agent "My Bot" --model gpt-4o
 
# Create an agent with a system prompt
puppyone conn add agent "Analyst" --model gpt-4o --system-prompt "You are a data analyst"

MCP endpoints

Create an MCP endpoint so clients such as Cursor and Claude Desktop can access your data directly:

puppyone conn add mcp "My Endpoint"

After creation, use conn key to retrieve the access key and configure it in the client.

Sandboxes

Create an isolated code execution environment:

# Default sandbox
puppyone conn add sandbox "Runner"
 
# Explicit E2B sandbox
puppyone conn add sandbox "Runner" --type e2b

Local folder sync

Set up two-way sync between a local directory and the cloud Content Node tree:

puppyone conn add folder ~/workspace --name "Dev Sync"

Inspect and manage connections

List all connections

# List everything
puppyone conn ls
 
# Filter by provider
puppyone conn ls --provider notion
 
# Filter by status
puppyone conn ls --status active
 
# JSON output
puppyone conn ls --json

Show connection details

puppyone conn info conn_abc123

Pause and resume

# Pause a data-source sync
puppyone conn pause conn_abc123
 
# Resume it
puppyone conn resume conn_abc123

Delete a connection

puppyone conn rm conn_abc123

Access key management

Agents, MCP endpoints, and sandboxes all authenticate with access keys.

# Show the access key
puppyone conn key conn_abc123
 
# Regenerate the access key
puppyone conn key conn_abc123 --regenerate

Sync and execution

Trigger sync manually

For data-source connections, run one sync immediately:

puppyone conn refresh conn_abc123

Set trigger mode

Configure how a connection is triggered:

puppyone conn trigger conn_abc123 manual
puppyone conn trigger conn_abc123 scheduled

Run once

Execute the connection logic directly:

puppyone conn run conn_abc123

View execution logs

puppyone conn logs conn_abc123

Update configuration

Modify the configuration of an existing connection:

# Update one setting
puppyone conn update conn_abc123 --set name="New Name"
 
# Update multiple settings with JSON
puppyone conn update conn_abc123 --config '{"folder": "/new-path", "mode": "incremental"}'

Show config fields

Inspect which config fields a given connection type supports:

puppyone conn schema notion
puppyone conn schema agent

Practical workflows

Set up multiple data sources for a project

# Connect Notion docs
puppyone conn add notion https://notion.so/wiki-page --folder /wiki
 
# Connect a GitHub repo
puppyone conn add github https://github.com/org/docs --folder /github
 
# Connect Gmail
puppyone conn add gmail
 
# Inspect connection health
puppyone conn ls

Create and use an MCP endpoint

# Create the endpoint
puppyone conn add mcp "Cursor Endpoint"
 
# Get the access key
puppyone conn key <id>
 
# Use that key in Cursor MCP configuration

Create and manage a sandbox

# Create a sandbox
puppyone conn add sandbox "Code Runner" --type e2b
 
# Inspect details
puppyone conn info <id>
 
# Get the access key
puppyone conn key <id>
 
# Trigger one execution
puppyone conn run <id>
 
# View execution logs
puppyone conn logs <id>

Manage many connections in bulk

# List paused connections
puppyone conn ls --status paused --json
 
# Resume all paused connections from a script
for id in $(puppyone conn ls --status paused --json | jq -r '.[].id'); do
  puppyone conn resume "$id"
done

File System Commands Local Folder Sync