openclaw-notion-skill（OpenClaw Notion 技能）

Notion Integration

Connect your Notion workspace to OpenClaw for seamless knowledge management and project tracking.

When to Use This Skill

Use Notion when the user wants to:

• Add items to a database (backlog, todos, tracking)
• Create new pages in a database or as children of existing pages
• Query/search their Notion workspace for information
• Update existing pages (status, notes, properties)
• Read page content or database entries

Setup

1. Create Notion Integration

1. Go to notion.so/my-integrations
2. Click New integration
3. Name it (e.g., "OpenClaw")
4. Select your workspace
5. Copy the Internal Integration Token (starts with secret_)
6. Save this token securely in OpenClaw config or environment: NOTION_TOKEN=secret_...

2. Share Pages with Integration

Important: Notion integrations have NO access by default. You must explicitly share:

1. Go to any page or database in Notion
2. Click Share → Add connections
3. Select your "OpenClaw" integration
4. The skill can now read/write to that specific page/database

3. Get Database/Page IDs

From URL:

• Database: https://www.notion.so/workspace/XXXXXXXX?v=... → ID is XXXXXXXX (32 chars)
• Page: https://www.notion.so/workspace/XXXXXXXX → ID is XXXXXXXX

Note: Remove hyphens when using IDs. Use the 32-character string.

Core Operations

Query Database

Retrieve entries from any database you've shared.

// Using the Notion skill via exec
await exec({
  command: `node ~/.agents/skills/notion/notion-cli.js query-database ${databaseId}`
});

// With filters (example: status = "In Progress")
await exec({
  command: `node ~/.agents/skills/notion/notion-cli.js query-database ${databaseId} --filter '{"property":"Status","select":{"equals":"In Progress"}}'`
});

Returns: Array of pages with properties as configured in your database.

Add Database Entry

Create a new row in a database.

// Add entry with multiple properties
await exec({
  command: `node ~/.agents/skills/notion/notion-cli.js add-entry ${databaseId} \
    --title "My New Content Idea" \
    --properties '${JSON.stringify({
      "Status": { "select": { "name": "Idea" } },
      "Platform": { "multi_select": [{ "name": "X/Twitter" }] },
      "Tags": { "multi_select": [{ "name": "3D Printing" }, { "name": "AI" }] },
      "Priority": { "select": { "name": "High" } }
    })}'`
});

Get Page Content

Read the content of any page (including database entries).

await exec({
  command: `node ~/.agents/skills/notion/notion-cli.js get-page ${pageId}`
});

Returns: Page title, properties, and block content (text, headings, lists, etc.).

Update Page

Modify properties or append content to an existing page.

// Update properties
await exec({
  command: `node ~/.agents/skills/notion/notion-cli.js update-page ${pageId} \
    --properties '${JSON.stringify({
      "Status": { "select": { "name": "In Progress" } }
    })}'`
});

// Append content blocks
await exec({
  command: `node ~/.agents/skills/notion/notion-cli.js append-body ${pageId} \
    --text "Research Notes" --type h2`
});

Search Notion

Find pages across your shared workspace.

await exec({
  command: `node ~/.agents/skills/notion/notion-cli.js search "content ideas"`
});

Common Use Cases

Content Pipeline (Content Creator Workflow)

Database Structure:

• Title (title)
• Status (select: Idea → Draft → Scheduled → Posted)
• Platform (multi_select: X/Twitter, YouTube, MakerWorld, Blog)
• Publish Date (date)
• Tags (multi_select)
• Draft Content (rich_text)

OpenClaw Integration:

// Research scout adds findings to Notion
await exec({
  command: `node ~/.agents/skills/notion/notion-cli.js add-entry ${contentDbId} \
    --title "New 3D Print Technique" \
    --properties '${JSON.stringify({
      "Status": { "select": { "name": "Idea" } },
      "Platform": { "multi_select": [{ "name": "YouTube" }] },
      "Tags": { "multi_select": [{ "name": "3D Printing" }] }
    })}'`
});

// Later: Update when drafting
await exec({
  command: `node ~/.agents/skills/notion/notion-cli.js update-page ${entryId} \
    --properties '${JSON.stringify({
      "Status": { "select": { "name": "Draft" } },
      "Draft Content": { "rich_text": [{ "text": { "content": "Draft text here..." } }] }
    })}'`
});

Project Management (Solo Entrepreneur)

Database Structure:

• Name (title)
• Status (select: Not Started → In Progress → Blocked → Done)
• Priority (select: Low → Medium → High → Critical)
• Due Date (date)
• Estimated Hours (number)
• Actual Hours (number)
• Links (url)
• Notes (rich_text)

Weekly Review Integration:

// Query all "In Progress" projects
await exec({
  command: `node ~/.agents/skills/notion/notion-cli.js query-database ${projectsDbId} --filter '{"property":"Status","select":{"equals":"In Progress"}}'`
});

Customer/Quote CRM (3D Printing Business)

Database Structure:

• Customer Name (title)
• Status (select: Lead → Quote Sent → Ordered → Printing → Shipped)
• Email (email)
• Quote Value (number)
• Filament Type (select)
• Due Date (date)
• Shopify Order ID (rich_text)

Shopify Integration:

// New order → create CRM entry
await exec({
  command: `node ~/.agents/skills/notion/notion-cli.js add-entry ${crmDbId} \
    --title "${customerName}" \
    --properties '${JSON.stringify({
      "Status": { "select": { "name": "Ordered" } },
      "Email": { "email": customerEmail },
      "Shopify Order ID": { "rich_text": [{ "text": { "content": orderId } }] }
    })}'`
});

Knowledge Base (Wiki Replacement for MEMORY.md)

Structure: Hub page with nested pages:

• 🏠 Home (shared with integration)

- SOPs - Troubleshooting - Design Patterns - Resource Links

Query for quick reference:

// Search for "stringing" to find 3D print troubleshooting
await exec({
  command: `node ~/.agents/skills/notion/notion-cli.js search "stringing"`
});

Property Types Reference

When creating/updating database entries, use these property value formats:

// Title (always required for new pages)
{ "title": [{ "text": { "content": "Page Title" } }] }

// Select (single choice)
{ "select": { "name": "Option Name" } }

// Multi-select (multiple choices)
{ "multi_select": [{ "name": "Tag 1" }, { "name": "Tag 2" }] }

// Status (for new Status property type)
{ "status": { "name": "In progress" } }

// Text / Rich text
{ "rich_text": [{ "text": { "content": "Your text here" } }] }

// Number
{ "number": 42 }

// Date
{ "date": { "start": "2026-02-15" } }
{ "date": { "start": "2026-02-15T10:00:00", "end": "2026-02-15T12:00:00" } }

// Checkbox
{ "checkbox": true }

// Email
{ "email": "user@example.com" }

// URL
{ "url": "https://example.com" }

// Phone
{ "phone_number": "+1-555-123-4567" }

// Relation (link to another database entry)
{ "relation": [{ "id": "related-page-id-32chars" }] }

Security & Permissions

Critical Security Model:

• ✅ Integration ONLY sees pages you explicitly share
• ✅ You control access per page/database
• ✅ Token stored securely in ~/.openclaw/.env (never in code)
• ❌ Never commit NOTION_TOKEN to git
• ❌ Integration cannot access private teamspaces or other users' private pages

Best Practices:

1. Use a dedicated integration (don't reuse personal integrations)
2. Share minimum necessary pages (granular > broad)
3. Rotate token if compromised via Notion integration settings
4. Review shared connections periodically

Environment Setup

Add to ~/.openclaw/.env:

NOTION_TOKEN=secret_xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

Or set per-command:

NOTION_TOKEN=secret_xxx node notion-cli.js ...

Error Handling

Common errors and fixes:

Quick Install (One Command)

cd ~/.agents/skills/notion
./install.sh

Manual install (if above fails):

cd ~/.agents/skills/notion
npm install

That's it! No build step required for the standalone version.

Quick Test

# After setting NOTION_TOKEN in ~/.openclaw/.env
node notion-cli.js test

Smart ID Resolution

Reference entries by Notion auto-ID (e.g., #3) or direct UUID.

By Notion ID (Recommended for Manual Use)

Use the number you see in your database's ID column:

# Get entry #3
node notion-cli.js get-page '#3' DATABASE_ID

# Add content to entry #3
node notion-cli.js append-body '#3' --database DATABASE_ID \
  --text "Research notes" --type h2

# Add bullet to entry #3
node notion-cli.js append-body '#3' --database DATABASE_ID \
  --text "Key finding" --type bullet

By Direct UUID (For Automation)

# Using full UUID from Notion URL
node notion-cli.js get-page 2fb3e4ac...
node notion-cli.js append-body 2fb3e4ac... \
  --text "Content" --type paragraph

Auto-detection: Starts with # = Notion ID lookup. 32-char hex = Direct UUID.

Pro Tip: Add an ID property (type: unique ID) to auto-number entries as #1, #2, #3...

Page Body Editing

Add rich content to page bodies, not just properties.

Append Content Blocks

# Add heading
node notion-cli.js append-body PAGE_ID --text "Research Summary" --type h2

# Add paragraph (default)
node notion-cli.js append-body PAGE_ID --text "Detailed findings go here..."

# Add bullet list item
node notion-cli.js append-body PAGE_ID --text "First key finding" --type bullet

# Add numbered list item
node notion-cli.js append-body PAGE_ID --text "Step one description" --type numbered

# Add TODO checkbox
node notion-cli.js append-body PAGE_ID --text "Create video script" --type todo

# Add quote
node notion-cli.js append-body PAGE_ID --text "Important quote from source" --type quote

# Add code block
node notion-cli.js append-body PAGE_ID --text "const result = optimizeSupports();" --type code --lang javascript

Supported Block Types

Get Page with Body Content

# Get full page including formatted body
node notion-cli.js get-page PAGE_ID

Returns:

• Page properties
• Formatted body blocks (type + content preview)
• Block count

Advanced: Raw JSON Blocks

For complex layouts, use raw Notion block JSON:

node notion-cli.js append-body PAGE_ID --blocks '[
  {"object":"block","type":"heading_2","heading_2":{"rich_text":[{"text":{"content":"Research Notes"}}]}},
  {"object":"block","type":"bulleted_list_item","bulleted_list_item":{"rich_text":[{"text":{"content":"Finding 1"}}]}},
  {"object":"block","type":"code","code":{"rich_text":[{"text":{"content":"console.log(1)"}}],"language":"javascript"}}
]'

Advanced: Webhook Sync

For bidirectional sync (Notion changes → OpenClaw):

1. Set up Notion webhook integration (requires Notion partner account)
2. Configure webhook endpoint to your OpenClaw Gateway
3. Skill processes incoming webhooks and updates memory files

See references/webhooks.md for implementation details.

Need help? Check your Notion integration settings at https://www.notion.so/my-integrations

Using in OpenClaw

Quick Setup

# 1. Install
cd ~/.agents/skills/notion
npm install

# 2. Configure token
echo "NOTION_TOKEN=secret_xxxxxxxxxx" >> ~/.openclaw/.env

# 3. Test connection
node notion-cli.js test

From OpenClaw Agent

// Query database
await exec({
  command: `node ~/.agents/skills/notion/notion-cli.js query-database YOUR_DB_ID`
});

// Add entry
await exec({
  command: `node ~/.agents/skills/notion/notion-cli.js add-entry YOUR_DB_ID \\
    --title "New Content Idea" \\
    --properties '{"Status":{"select":{"name":"Idea"}}}'`
});

// Search
await exec({
  command: `node ~/.agents/skills/notion/notion-cli.js search "tree support"`
});

Cron Job Usage

Update your Research Topic Scout to push to Notion:

"message": "Research trends and add to Notion: 
  node ~/.agents/skills/notion/notion-cli.js add-entry DB_ID 
    --title '<title>' 
    --properties '{...,\"Platform\":{\"multi_select\":[{\"name\":\"X\"}]}}'"