Advanced 📚
Advanced 📚
Managing messenger inboxes
Manage all your messages across Slack, LinkedIn, X (Twitter), WhatsApp, Instagram, and more in one inbox—then let Alice summarize, prioritize, and draft responses in your personal voice.
What You'll Need
Beeper app (beeper.com) - Free version works perfectly
Alice desktop app with MCP support enabled
A smart AI model - Claude Sonnet 3.5 or GPT-4 recommended (avoid GPT-3.5 for MCP tasks)
Step 1: Set Up Beeper
Install and Configure Beeper
Download Beeper from beeper.com
Connect your messaging accounts:
Open Beeper → Preferences → Add Account
Connect: Slack, X (Twitter), LinkedIn, Instagram, WhatsApp, etc.
All messages now appear in one unified inbox
Step 2: Enable Beeper MCP Server
In Beeper:
Open Beeper → Preferences → Developers tab
Find "Beeper MCP" section (disabled by default)
Click "Start" to launch the MCP server
Note the port number displayed (you'll need this)
Optional: Enable "Start on launch" for automatic startup
Step 3: Generate Authentication Token
Create API Connection:
In Beeper Developers tab, find "Approved Connections"
Click "+ Create New Connection"
Copy the generated API key (bearer token)
Keep this handy for the next step
Step 4: Configure Alice
Add Beeper MCP Server:
Open Alice → ⌘K (or ⌘, for Settings) → MCP section
Click "Add Server" button
Configure the connection:
Installation Settings:
Type: Switch to "Stream/HTTP and SSE"
Address: Copy the exact address from Beeper's Developer tab
(Example:http://localhost:PORT/mcp- use YOUR port number)
Advanced Configuration:
Click the arrow to reveal "HTTP Headers"
Add header:
Field 1:
AuthorizationField 2:
Bearer YOUR_API_KEY_HERE
(Paste the token from Step 3 - no extra spaces or quotes)
Assign to an Assistant:
Select which Alice assistant(s) can use Beeper
Toggle "Enable" for this connection
Step 5: Verify Connection
Test the Setup:
Make sure Beeper's MCP server is running (check Developer tab)
In Alice MCP settings, disable then re-enable the Beeper connection
Status should show "Connected"
Click to see 10 available tools:
Get accounts
Archive chats
Search messages
Send messages
Open chats
And more...
Step 6: Personalize Your Assistant
Customize Response Style:
Go to your assistant's Master Prompt
Add instructions for how to handle messages:
Tone of voice: Professional, casual, friendly, etc.
Response rules: "Decline conference invitations politely"
Examples: "When someone asks for X, I usually reply with..."
Pro Tip: Create Skills
Create reusable prompts like:
"Kindly refuse invitation"
"Draft professional follow-up"
"Summarize and prioritize messages"
Use Skills + Assistant + Beeper together for instant, personalized message handling.
Step 7: Start Using Alice with Beeper
Example Commands:
Simple queries:
"What are my unread messages on Beeper?"
"Show me new messages"
Advanced automation:
"Summarize my latest messages and prioritize them"
"Categorize messages into work and personal"
"Draft polite responses to the top 3 work messages"
"Send a reply to [person] saying I'll follow up tomorrow"
Important Notes:
Use smart models (Claude Sonnet 3.5 or GPT-4) for best results
Experiment with prompts - be specific about what you want
Mention "Beeper" explicitly if Alice doesn't automatically use the MCP connection
What You Can Do Now
One inbox for Slack, WhatsApp, LinkedIn, X, Instagram, and more
AI-powered summaries of unread messages
Automated prioritization by urgency or category
Draft responses in YOUR voice based on your assistant's training
Send messages directly from Alice with a single command
Save hours on message management every week
Pro Tips
Auto-start everything: Enable both Beeper and its MCP server to launch on system startup
Model selection matters: Use Claude Sonnet 3.5 for complex message workflows
Read tool descriptions: Review available Beeper tools in Alice's MCP settings to understand capabilities
Link to assistants: MCP connections only work when assigned to at least one assistant
Combine with Skills: Create reusable message templates for common scenarios
Troubleshooting
Connection Failed?
Check that Beeper's MCP server is running (Developer tab)
Verify the port number matches between Beeper and Alice
Ensure bearer token is copied correctly (no extra spaces)
"MCP not active" message?
Make sure at least one assistant has the Beeper connection enabled
Check that "Tools" are enabled in your assistant settings
AI not using Beeper?
Explicitly mention "Beeper" in your query
Switch to a smarter model (Claude Sonnet 3.5 recommended)
Verify tools are enabled for your active assistant
What's Next?
Now that you have unified message management with AI automation, you can:
Build custom workflows for different message types
Create assistant personas for work vs personal communication
Automate daily message triage with scheduled prompts
Explore other MCP integrations to connect more tools to Alice
Happy automating!
For more tutorials and tips, subscribe to our YouTube channel and check out the Alice documentation.
What You'll Need
Beeper app (beeper.com) - Free version works perfectly
Alice desktop app with MCP support enabled
A smart AI model - Claude Sonnet 3.5 or GPT-4 recommended (avoid GPT-3.5 for MCP tasks)
Step 1: Set Up Beeper
Install and Configure Beeper
Download Beeper from beeper.com
Connect your messaging accounts:
Open Beeper → Preferences → Add Account
Connect: Slack, X (Twitter), LinkedIn, Instagram, WhatsApp, etc.
All messages now appear in one unified inbox
Step 2: Enable Beeper MCP Server
In Beeper:
Open Beeper → Preferences → Developers tab
Find "Beeper MCP" section (disabled by default)
Click "Start" to launch the MCP server
Note the port number displayed (you'll need this)
Optional: Enable "Start on launch" for automatic startup
Step 3: Generate Authentication Token
Create API Connection:
In Beeper Developers tab, find "Approved Connections"
Click "+ Create New Connection"
Copy the generated API key (bearer token)
Keep this handy for the next step
Step 4: Configure Alice
Add Beeper MCP Server:
Open Alice → ⌘K (or ⌘, for Settings) → MCP section
Click "Add Server" button
Configure the connection:
Installation Settings:
Type: Switch to "Stream/HTTP and SSE"
Address: Copy the exact address from Beeper's Developer tab
(Example:http://localhost:PORT/mcp- use YOUR port number)
Advanced Configuration:
Click the arrow to reveal "HTTP Headers"
Add header:
Field 1:
AuthorizationField 2:
Bearer YOUR_API_KEY_HERE
(Paste the token from Step 3 - no extra spaces or quotes)
Assign to an Assistant:
Select which Alice assistant(s) can use Beeper
Toggle "Enable" for this connection
Step 5: Verify Connection
Test the Setup:
Make sure Beeper's MCP server is running (check Developer tab)
In Alice MCP settings, disable then re-enable the Beeper connection
Status should show "Connected"
Click to see 10 available tools:
Get accounts
Archive chats
Search messages
Send messages
Open chats
And more...
Step 6: Personalize Your Assistant
Customize Response Style:
Go to your assistant's Master Prompt
Add instructions for how to handle messages:
Tone of voice: Professional, casual, friendly, etc.
Response rules: "Decline conference invitations politely"
Examples: "When someone asks for X, I usually reply with..."
Pro Tip: Create Skills
Create reusable prompts like:
"Kindly refuse invitation"
"Draft professional follow-up"
"Summarize and prioritize messages"
Use Skills + Assistant + Beeper together for instant, personalized message handling.
Step 7: Start Using Alice with Beeper
Example Commands:
Simple queries:
"What are my unread messages on Beeper?"
"Show me new messages"
Advanced automation:
"Summarize my latest messages and prioritize them"
"Categorize messages into work and personal"
"Draft polite responses to the top 3 work messages"
"Send a reply to [person] saying I'll follow up tomorrow"
Important Notes:
Use smart models (Claude Sonnet 3.5 or GPT-4) for best results
Experiment with prompts - be specific about what you want
Mention "Beeper" explicitly if Alice doesn't automatically use the MCP connection
What You Can Do Now
One inbox for Slack, WhatsApp, LinkedIn, X, Instagram, and more
AI-powered summaries of unread messages
Automated prioritization by urgency or category
Draft responses in YOUR voice based on your assistant's training
Send messages directly from Alice with a single command
Save hours on message management every week
Pro Tips
Auto-start everything: Enable both Beeper and its MCP server to launch on system startup
Model selection matters: Use Claude Sonnet 3.5 for complex message workflows
Read tool descriptions: Review available Beeper tools in Alice's MCP settings to understand capabilities
Link to assistants: MCP connections only work when assigned to at least one assistant
Combine with Skills: Create reusable message templates for common scenarios
Troubleshooting
Connection Failed?
Check that Beeper's MCP server is running (Developer tab)
Verify the port number matches between Beeper and Alice
Ensure bearer token is copied correctly (no extra spaces)
"MCP not active" message?
Make sure at least one assistant has the Beeper connection enabled
Check that "Tools" are enabled in your assistant settings
AI not using Beeper?
Explicitly mention "Beeper" in your query
Switch to a smarter model (Claude Sonnet 3.5 recommended)
Verify tools are enabled for your active assistant
What's Next?
Now that you have unified message management with AI automation, you can:
Build custom workflows for different message types
Create assistant personas for work vs personal communication
Automate daily message triage with scheduled prompts
Explore other MCP integrations to connect more tools to Alice
Happy automating!
For more tutorials and tips, subscribe to our YouTube channel and check out the Alice documentation.
Advanced 📚
Managing messenger inboxes
Manage all your messages across Slack, LinkedIn, X (Twitter), WhatsApp, Instagram, and more in one inbox—then let Alice summarize, prioritize, and draft responses in your personal voice.
What You'll Need
Beeper app (beeper.com) - Free version works perfectly
Alice desktop app with MCP support enabled
A smart AI model - Claude Sonnet 3.5 or GPT-4 recommended (avoid GPT-3.5 for MCP tasks)
Step 1: Set Up Beeper
Install and Configure Beeper
Download Beeper from beeper.com
Connect your messaging accounts:
Open Beeper → Preferences → Add Account
Connect: Slack, X (Twitter), LinkedIn, Instagram, WhatsApp, etc.
All messages now appear in one unified inbox
Step 2: Enable Beeper MCP Server
In Beeper:
Open Beeper → Preferences → Developers tab
Find "Beeper MCP" section (disabled by default)
Click "Start" to launch the MCP server
Note the port number displayed (you'll need this)
Optional: Enable "Start on launch" for automatic startup
Step 3: Generate Authentication Token
Create API Connection:
In Beeper Developers tab, find "Approved Connections"
Click "+ Create New Connection"
Copy the generated API key (bearer token)
Keep this handy for the next step
Step 4: Configure Alice
Add Beeper MCP Server:
Open Alice → ⌘K (or ⌘, for Settings) → MCP section
Click "Add Server" button
Configure the connection:
Installation Settings:
Type: Switch to "Stream/HTTP and SSE"
Address: Copy the exact address from Beeper's Developer tab
(Example:http://localhost:PORT/mcp- use YOUR port number)
Advanced Configuration:
Click the arrow to reveal "HTTP Headers"
Add header:
Field 1:
AuthorizationField 2:
Bearer YOUR_API_KEY_HERE
(Paste the token from Step 3 - no extra spaces or quotes)
Assign to an Assistant:
Select which Alice assistant(s) can use Beeper
Toggle "Enable" for this connection
Step 5: Verify Connection
Test the Setup:
Make sure Beeper's MCP server is running (check Developer tab)
In Alice MCP settings, disable then re-enable the Beeper connection
Status should show "Connected"
Click to see 10 available tools:
Get accounts
Archive chats
Search messages
Send messages
Open chats
And more...
Step 6: Personalize Your Assistant
Customize Response Style:
Go to your assistant's Master Prompt
Add instructions for how to handle messages:
Tone of voice: Professional, casual, friendly, etc.
Response rules: "Decline conference invitations politely"
Examples: "When someone asks for X, I usually reply with..."
Pro Tip: Create Skills
Create reusable prompts like:
"Kindly refuse invitation"
"Draft professional follow-up"
"Summarize and prioritize messages"
Use Skills + Assistant + Beeper together for instant, personalized message handling.
Step 7: Start Using Alice with Beeper
Example Commands:
Simple queries:
"What are my unread messages on Beeper?"
"Show me new messages"
Advanced automation:
"Summarize my latest messages and prioritize them"
"Categorize messages into work and personal"
"Draft polite responses to the top 3 work messages"
"Send a reply to [person] saying I'll follow up tomorrow"
Important Notes:
Use smart models (Claude Sonnet 3.5 or GPT-4) for best results
Experiment with prompts - be specific about what you want
Mention "Beeper" explicitly if Alice doesn't automatically use the MCP connection
What You Can Do Now
One inbox for Slack, WhatsApp, LinkedIn, X, Instagram, and more
AI-powered summaries of unread messages
Automated prioritization by urgency or category
Draft responses in YOUR voice based on your assistant's training
Send messages directly from Alice with a single command
Save hours on message management every week
Pro Tips
Auto-start everything: Enable both Beeper and its MCP server to launch on system startup
Model selection matters: Use Claude Sonnet 3.5 for complex message workflows
Read tool descriptions: Review available Beeper tools in Alice's MCP settings to understand capabilities
Link to assistants: MCP connections only work when assigned to at least one assistant
Combine with Skills: Create reusable message templates for common scenarios
Troubleshooting
Connection Failed?
Check that Beeper's MCP server is running (Developer tab)
Verify the port number matches between Beeper and Alice
Ensure bearer token is copied correctly (no extra spaces)
"MCP not active" message?
Make sure at least one assistant has the Beeper connection enabled
Check that "Tools" are enabled in your assistant settings
AI not using Beeper?
Explicitly mention "Beeper" in your query
Switch to a smarter model (Claude Sonnet 3.5 recommended)
Verify tools are enabled for your active assistant
What's Next?
Now that you have unified message management with AI automation, you can:
Build custom workflows for different message types
Create assistant personas for work vs personal communication
Automate daily message triage with scheduled prompts
Explore other MCP integrations to connect more tools to Alice
Happy automating!
For more tutorials and tips, subscribe to our YouTube channel and check out the Alice documentation.