Introduction to the Sales Plugin
The Sales plugin is one of the most powerful tools for sales professionals in Claude Cowork. It combines AI intelligence with your existing sales tools to streamline prospect research, call preparation, pipeline management, and outreach.
What the Sales Plugin Can Do
- Prospect Research: Automatically gather company information, news, and key contacts
- Call Preparation: Generate comprehensive call prep documents before meetings
- Pipeline Review: Analyze your deals and identify risks or opportunities
- Outreach Drafting: Create personalized emails and messages at scale
- Battlecards: Access competitive intelligence when you need it
Supported Integrations
The Sales plugin connects to:
- CRM: HubSpot, Close
- Data Enrichment: Clay, ZoomInfo
- Communication: Slack, Microsoft 365
- Documentation: Notion, Jira
- Call Intelligence: Fireflies
Installation
Step 1: Install the Plugin
claude plugin install sales@knowledge-work-plugins
Or in Claude Cowork desktop:
- Go to Plugins → Browse
- Find "Sales" and click Install
Step 2: Configure Your CRM
Edit the .mcp.json file to add your HubSpot or Close credentials:
{
"mcpServers": {
"hubspot": {
"command": "npx",
"args": ["-y", "@anthropic/mcp-server-hubspot"],
"env": {
"HUBSPOT_API_KEY": "your-api-key"
}
}
}
}
Step 3: Add Your Sales Process
Customize the skills files to reflect your sales methodology (MEDDIC, BANT, etc.).
Key Slash Commands
/sales:call-prep
Prepare for an upcoming call with comprehensive research:
/sales:call-prep Acme Corp - John Smith
Output includes:
- Company overview and recent news
- Contact's role and LinkedIn summary
- Previous interactions from your CRM
- Suggested talking points
- Potential objections and responses
/sales:research-company
Deep dive into a target company:
/sales:research-company Acme Corp
Output includes:
- Company size, funding, and financials
- Key decision makers
- Tech stack (if available)
- Recent news and announcements
- Competitors
/sales:pipeline-review
Analyze your current pipeline:
/sales:pipeline-review Q1 2026
Output includes:
- Deals at risk (stalled, no recent activity)
- Upcoming close dates
- Forecast analysis
- Recommended actions
/sales:draft-outreach
Create personalized outreach:
/sales:draft-outreach John Smith at Acme Corp - cold email introducing our product
Real-World Workflows
Morning Pipeline Review
Start your day with:
Review my deals closing this month. Flag any at risk and suggest next steps for each.
Pre-Meeting Prep
Before every call:
/sales:call-prep [Company] - [Contact Name]
Give me 3 discovery questions based on their recent news.
Weekly Account Planning
End of week review:
Analyze my top 10 accounts. Which ones need more attention?
What's the best approach for each?
Batch Outreach
For prospecting:
I have these 20 contacts from the AI conference. Draft personalized
follow-up emails for each based on their company and role.
Customization Tips
Add Your Battlecards
Create a battlecards.md file in the skills folder with competitive intel:
## vs Competitor A
- Their weakness: Limited integrations
- Our strength: 50+ native integrations
- Key question to ask: "How important is connecting to your existing tools?"
Define Your Sales Process
Add your methodology to the skills:
## Our Sales Process
1. Discovery: Understand pain points using SPIN questions
2. Demo: Focus on top 3 use cases
3. Proposal: Always include ROI calculation
4. Negotiation: Max 10% discount without VP approval
Set Up Triggers
Configure automatic research for new leads in your CRM:
When a new lead is added to HubSpot, automatically run
/sales:research-company and attach findings to the contact record.
Best Practices
- Keep CRM updated: The plugin's insights are only as good as your data
- Customize for your industry: Add industry-specific knowledge to skills
- Review AI suggestions: Always personalize generated outreach
- Iterate on prompts: Save effective prompts as custom commands
Security Considerations
- API keys are stored locally in your
.mcp.json - Customer data passes through Anthropic's API (review your data policies)
- Don't include sensitive financial terms in prompts