Skip to main content
MCP

Top productivity MCPs for Claude Code

The best MCPs for productivity with Claude Code: Gmail, Slack, Google Calendar, Figma, and Lighthouse. Installation, configuration, and use cases.

MCPs that transform your daily workflow

Productivity MCPs connect Claude Code to your everyday tools: email, messaging, calendar, design, and web auditing. They turn Claude Code into a personal assistant that can take action in your real professional environment.


Gmail: Email management

Read, search, and manage your emails from Claude Code. Create drafts, analyze conversations, and extract key information from your inbox.

JSON configuration (generic example)

{
"mcpServers": {
"gmail": {
"command": "npx",
"args": ["-y", "gmail-mcp-package-name"],
"env": {
"GMAIL_OAUTH_CLIENT_ID": "your-client-id",
"GMAIL_OAUTH_CLIENT_SECRET": "your-client-secret",
"GMAIL_OAUTH_REFRESH_TOKEN": "your-refresh-token"
}
}
}
}

Use cases

  • Smart triage: "Show me this week's unread emails and summarize each one in a sentence"
  • Advanced search: "Find all emails from @client.com about project X"
  • Automatic drafts: "Draft a professional reply to this email addressing the points raised"
  • Information extraction: "Extract all meeting dates mentioned in my emails this week"

Slack: Team messaging

Interact with Slack directly from Claude Code. Read and send messages, manage channels, and search through your team's history.

Installation

# Slack requires a Bot Token
claude mcp add slack -e SLACK_BOT_TOKEN=xoxb-your-token -- npx -y @modelcontextprotocol/server-slack

JSON configuration

{
"mcpServers": {
"slack": {
"command": "npx",
"args": ["-y", "@modelcontextprotocol/server-slack"],
"env": {
"SLACK_BOT_TOKEN": "xoxb-your-bot-token",
"SLACK_TEAM_ID": "your-team-id"
}
}
}
}

Use cases

  • Team monitoring: "Summarize the important messages from the #engineering channel this week"
  • Automatic notifications: "Send a message to #releases announcing the new version"
  • Contextual search: "Find the conversation where we decided on the database architecture"

Google Calendar: Schedule management

View and manage your Google Calendar directly from Claude Code. See your events, find open time slots, and plan meetings.

JSON configuration (generic example)

{
"mcpServers": {
"google-calendar": {
"command": "npx",
"args": ["-y", "calendar-mcp-package-name"],
"env": {
"GOOGLE_OAUTH_CLIENT_ID": "your-client-id",
"GOOGLE_OAUTH_CLIENT_SECRET": "your-client-secret",
"GOOGLE_OAUTH_REFRESH_TOKEN": "your-refresh-token"
}
}
}
}

Use cases

  • Daily overview: "Show me tomorrow's schedule with details for each meeting"
  • Smart scheduling: "Find a free one-hour slot this week for a code review session"
  • Meeting prep: "Look at my next meeting and prepare a context summary for me"

Figma: Design and mockups

Access your Figma designs directly from Claude Code. Export components, analyze mockups, and integrate design into your development workflow.

JSON configuration (generic example)

{
"mcpServers": {
"figma": {
"command": "npx",
"args": ["-y", "figma-mcp-package-name"],
"env": {
"FIGMA_ACCESS_TOKEN": "figd_your-figma-token"
}
}
}
}

Use cases

  • Analyze a design: "Look at this Figma mockup and describe the components, colors, and spacing used"
  • Generate code from a design: "Turn this Figma component into a React component with Tailwind CSS"
  • Compare implementation to mockup: "Compare my current page with the Figma mockup and list the differences"

Lighthouse: Web performance auditing

Run Lighthouse audits directly from Claude Code. Analyze performance, accessibility, SEO, and best practices for any website.

Via the Playwright MCP (recommended)

{
"mcpServers": {
"playwright": {
"command": "npx",
"args": ["-y", "@playwright/mcp"]
}
}
}

Then in Claude Code:

"Run a Lighthouse audit on https://my-site.com and give me a summary of the scores"

Use cases

  • Full audit: "Run a Lighthouse audit on my homepage and give me the performance, accessibility, and SEO scores"
  • Targeted optimization: "Analyze my page's performance and suggest concrete fixes"
  • Improvement tracking: "Compare the Lighthouse scores before and after my changes"

Summary

MCPVerified packageAuth requiredBest for
GmailCommunity (check npm)Google OAuthEmail triage, search, drafts
Slack@modelcontextprotocol/server-slackBot TokenTeam communication, monitoring
Google CalendarCommunity (check npm)Google OAuthPlanning, organization
FigmaOfficial Figma (check developer.figma.com)Personal tokenDesign-to-code, mockup analysis
LighthouseVia Playwright MCP (@playwright/mcp)NoneWeb audits, performance optimization

Limitations and gotchas

Before plugging these MCPs into production, keep their real limitations in mind.

API quotas and rate limiting

  • Gmail: Google's API enforces daily quotas (250 requests/second by default). Heavy use of automatic triage can hit them.
  • Slack: rate limits vary by endpoint type (Tier 1 through 4). Sending messages is limited to roughly 1 request per second.
  • Google Calendar: same Google quotas as Gmail. Reads are more permissive than writes.
  • Figma: the personal token has request limits. Large files take time to analyze.

Latency and reliability

Productivity MCPs depend on third-party APIs. If Slack is down, the Slack MCP is too. Plan for clear error messages and don't rely on these MCPs for real-time critical workflows.

Common failure cases

  • Expired token: Google OAuth tokens expire regularly. Make sure you have a valid refresh token configured.
  • Insufficient permissions: a Slack Bot without the chat:write scope won't be able to send messages, even if the MCP starts correctly.
  • API changes: third-party APIs evolve. A community MCP can become incompatible overnight.

Evaluating a third-party MCP

Before installing a community MCP, check these criteria:

CriterionWhat to look for
MaintenanceRecent commits, issues being addressed, active contributors
SecuritySource code audit, no excessive permissions, no suspicious dependencies
PerformanceReasonable startup time, no known memory leaks
DocumentationClear README, configuration examples, maintained changelog

Next steps