Dextego MCP serverYour sales coach and team reports, inside the AI you already use
Connect Dextego to Claude, ChatGPT or any MCP-ready assistant. Sellers learn what to work on and walk into every meeting prepared. Managers see exactly where to coach, and can have a team report in their inbox every Monday without lifting a finger.
Three priorities for this week
- Probing questions are the lowest-scoring skill on the team.
- Price objections came up in 1 of every 3 discovery calls.
- Maya and Daniel improved empathy the most. Use their calls as examples.
Illustrative example
What you get out of it
Less time digging through calls, more time coaching and selling.
For sellers
- Know what to work on nextAsk how your week went and get your strongest skill, your biggest gain and the one thing to fix.
- Walk into every meeting preparedWho's attending, what they care about and how to connect with them, before you join.
- Handle the objections you actually getSee the pushback buyers raise most on your calls, and how your best answers went.
- Watch your influence growTrack your Influence Score week by week, and see which skills are moving it.
For managers
- Coach the right thing this weekGet your team's top coaching priorities and skill gaps in one answer.
- Spot who needs help earlySee who's improving, who's stuck and why, before the pipeline review does.
- Spread what your best reps doPull your team's best calls and turn them into playbook updates and objection guides.
- Reports without the spreadsheetWeekly team performance and leadership reports, written for you from real call data.
Dextego + your other tools
Watch your AI actually do the work
Connect Dextego alongside your CRM, Google Drive, Gmail and Slack in Claude. One request, and your AI pulls what happened on your calls and gets it where it needs to go.
- Dextegolist_upcoming_meetings2 meetings, with buyer profiles
- HubSpotRead deal historyStage, last activity and open questions
- Google DriveFind the latest proposalPricing sent on the 12th
- SlackSend you a DMOne brief per meeting
✓ Brief sent to your Slack
- 10:00 VP Sales: lead with numbers, keep it short
- 14:30 Head of RevOps: wants a rollout plan
- Open question: security review timing
Illustrative example
Put it on autopilot: reports in your inbox every week
You don’t have to remember to ask. With scheduled tasks in Claude, a prompt runs on its own, every day, week or month, and the answer is waiting for you.
- 1
Connect Dextego to Claude
Add Dextego as a connector, plus Gmail or Outlook if you want results by email.
- 2
Create a scheduled task
Write the prompt once and pick when it runs: daily, weekly or monthly.
- 3
Get it in your inbox
Claude runs it on schedule with your latest Dextego data and sends you the result.
“Summarise my team's week in Dextego: coaching priorities, skill gaps and most improved reps. Email it to me.”
“How did my calls go this week, and what's the one thing I should work on next week? Email it to me.”
“Prep me for today's meetings: who's attending, what they care about and how to connect with them.”
“Write a leadership report on last month's objections, Influence Score trends and best calls.”
Scheduled tasks run from the Claude desktop app. Other assistants, such as ChatGPT, offer similar scheduling.
What is Dextego MCP?
Dextego MCP is Dextego’s Model Context Protocol server. It lets AI assistants such as Claude and ChatGPT securely read your sales conversation data from Dextego, the AI sales coaching platform that scores calls and coaches reps live.
Sales reps, managers and enablement teams use it to find the objections buyers raise most, spot team skill gaps, prepare for meetings and update sales playbooks, all without leaving the AI assistant they already use.
- What it is
- A remote MCP server (Model Context Protocol) for the Dextego AI sales coaching platform
- Works with
- Claude, ChatGPT, Claude Code, Cursor, VS Code and any MCP client with OAuth sign-in
- Tools
- 32 read-only tools, including prepared analyses of calls, objections and skill gaps
- Data
- Your calls, transcripts, scores, objections, upcoming meetings, knowledge base and coaching history
- Security
- OAuth sign-in with your Dextego account; the AI only sees what you can see; team data is admin-only
- Availability
- Included on paid Dextego plans
Dextego MCP
How the coaching loop works
Every call makes the next one better. Dextego scores the conversation, your AI assistant turns the pattern into enablement, and the new playbook coaches the rep live on the next call.
Works with
- Claude
- ChatGPT
- Claude Code
- Cursor
- VS Code
Dextego reviews every call
Each conversation is scored on gravitas, empathy, objection handling and probing questions, and every buyer objection is logged.
Illustrative example. Tap any step on the ring to jump to it.
Ask it anything about your sales conversations
Dextego’s prepared analyses do the heavy lifting, so your AI answers with figures Dextego has already computed from your real calls. Click a question to see a sample answer.
For sellers
Your own calls, scores and coaching.
“How did I do this week?”
Your week, Maya: 9 calls, Influence Score up 4 points
- Best call: Northwind Logistics discovery (8.4/10), your strongest rapport this month.
- Biggest gain: active listening, up from 6.1 to 7.3. It showed on the Brightline Health demo.
- Work on next: next steps. 4 of 9 calls, including Atlas Freight, ended without a booked follow-up.
Illustrative example
“What objections am I getting most often?”
Your top 3 objections from the last 30 days
- “It’s too expensive” came up in 7 calls. Your best answer was on the Kestrel Bank call.
- “We already use Gong” came up in 5 calls, mostly in discovery, most recently with Harbor Retail.
- “Not before Q1” came up in 3 calls. Tie the timing to what waiting costs them.
Illustrative example
“Analyse my last few calls and tell me what to work on.”
Across your last 5 calls, from Northwind to Atlas Freight, focus on two things
- Ask more probing questions. You asked 3 on average. Your Northwind call had 8, and it moved to a demo.
- Talk less in discovery. Your talk time on Atlas Freight was 64%. Aim for under 45%.
- Keep doing: clear recaps. Buyers confirmed your summary on every call.
Illustrative example
“How has my Influence Score changed over the last couple of months?”
Up from 62 to 71 over the last 8 weeks
- Steady climb since week 3, when you started booking next steps on the call.
- One dip in week 5: two short calls with Harbor Retail where the buyer barely spoke.
- Empathy contributed most to the gain (+1.8).
Illustrative example
“Prep me for my next meeting. Who's attending and what do they care about?”
Tomorrow 10:00, Kestrel Bank discovery call
- Attendees: Rachel Moore, VP of Sales, and Tom Nguyen, RevOps manager.
- What they care about: ramp time for 12 new hires and forecast accuracy.
- How to connect: Rachel is data-driven. Lead with numbers and keep it concise.
Illustrative example
For managers and enablement
Team analyses for organisation admins.
“What should I coach my team on this week?”
Three priorities for this week
- Probing questions are the lowest-scoring skill, especially for Liam and Sofia.
- Price objections came up in 1 of every 3 discovery calls, including Atlas Freight and Kestrel Bank.
- Maya and Daniel improved empathy the most. Use their calls as examples.
Illustrative example
“What objections are we hearing most often, and how should we answer them?”
Top objections across 142 team calls this month
- Budget (31%): sellers who reframed around the cost of waiting won the next step 2x as often. Daniel did this on Brightline Health.
- Already using Gong (22%): the best answers asked what the current tool misses, as Maya did with Northwind.
- Timing (15%): usually raised late, as on Harbor Retail. Surface it earlier in discovery.
Illustrative example
“What are the biggest skill gaps across my team?”
Largest gaps against your team's target scores
- Closing and next steps: 5.8 vs a target of 7.5, the widest gap. Liam has 3 deals with no next step booked.
- Probing questions: 6.2 vs 7.5, most visible with Sofia and Liam, who joined this quarter.
- Strength: rapport building is above target for 8 of 10 sellers.
Illustrative example
“Who on my team has improved the most recently, and in what skills?”
Most improved over the last 30 days
- Daniel Okafor: +9 Influence Score, driven by objection handling on Brightline Health.
- Maya Chen: +7, with the biggest jump in active listening.
- Sofia Alvarez: +5, now booking next steps on 80% of calls.
Illustrative example
“Show me examples of great selling from my team.”
3 standout calls worth sharing
- Maya, Northwind Logistics discovery (9.1): a textbook example of layered probing questions.
- Daniel, Brightline Health demo (8.8): turned a pricing objection into a pilot agreement.
- Sofia, Kestrel Bank follow-up (8.6): strong recap and a clear, mutual next step.
Illustrative example
Connect in a couple of minutes
In the Dextego app, open your account menu and choose Connect AI tools. Pick your assistant, paste the link, sign in, and you’re set. Dextego MCP is included on paid Dextego plans.
Open DextegoWhat your AI can see, and what it can’t
Read-only
It can't change or delete anything.
Your permissions
It only sees what you can see.
You approve access
Team data stays admin-only.
Frequently asked questions
Dextego MCP is the Model Context Protocol (MCP) server for Dextego, the AI sales coaching platform. It gives AI assistants like Claude, ChatGPT and Cursor secure, read-only access to your Dextego calls, transcripts, scores, buyer objections, upcoming meetings, knowledge base and coaching history, so you can ask questions about your sales conversations in plain language.
Yes. Dextego offers a remote MCP server with 32 read-only tools. It works with any AI tool that supports remote MCP servers over HTTP with OAuth sign-in, including Claude, ChatGPT, Claude Code, Cursor and VS Code.
MCP (Model Context Protocol) is an open standard that lets AI assistants use tools and data from other apps. Instead of copying and pasting call notes into a chat, the assistant fetches the data it needs from Dextego directly.
Sign in to the Dextego app, open your account menu and choose Connect AI tools. Copy your connector URL, add it as a custom connector in Claude or ChatGPT, then sign in with your Dextego account and approve the permissions. It takes about two minutes.
Sellers can ask things like “How did I do this week?” or “What objections am I getting most often?”. Managers can ask “What should I coach my team on this week?”, “What are the biggest skill gaps across my team?” or “Show me examples of great selling from my team.” The assistant can then turn the answers into objection handling guides, playbook updates and leadership reports.
Yes. Connect Dextego to Claude, along with an email connector such as Gmail or Outlook, then create a scheduled task, for example “Every Monday at 8:00, summarise my team’s week in Dextego and email it to me.” Claude runs the prompt on schedule with your latest Dextego data and sends you the result.
No. Every Dextego MCP tool is read-only. Your assistant can look things up, but it can’t create, edit or delete anything in Dextego.
Everyone can use the tools for their own calls, knowledge base and coaching. Team analyses, team members, performance reports and billing are only available to organisation admins. The AI only ever sees what your Dextego account can see.
Connecting AI tools is included on paid Dextego plans. Admins can upgrade from billing in the Dextego app, and plan details are on the Dextego pricing page.
