GitHub
GitHub MCP Connector for Databox
Connect GitHub to your AI Analyst through MCP. It reads the commits and pull requests behind your engineering metrics, explaining a spike or slowdown with the real change. GitHub's MCP server is open to every account today.
TL;DR
Get more context from your GitHub data
- Same MCP server, wired into Databox. The GitHub connector links GitHub's own MCP server to your AI Analyst, open to any GitHub account today as a Custom connector.
- More than the numbers: your AI Analyst reads the actual commits and pull requests behind your engineering metrics.
- Sharper answers: it can explain a spike in deploys or a slowdown in review time with the real change behind it, not just the count.
The basics
An MCP connector is not the same as an integration
GitHub has both. They do different jobs and work well together.
GitHub integration
- Pulls your GitHub metrics into Databox dashboards and reports - with historical data, custom metrics, and datasets you can build on.
GitHub MCP connector
- Gives your AI Analyst live access to GitHub's own MCP server, so it can explain metrics using the commits and pull requests behind them.
How to use it
Step 1
Open connector settings
Go to your AI Analyst settings → Connectors.
Step 2
Add a custom connector
Select Add custom connector, name it GitHub and paste the URL of GitHub's own MCP server.
Step 3
Authenticate with GitHub
Sign in with your GitHub account and set tool permissions per tool.
Step 4
Ask anything
Ask a question that touches GitHub - it now pulls the record behind the number automatically.
The Databox difference
Why connecting MCP through Databox is different
You can connect this same MCP server directly to any AI tool. Here’s what changes when you connect it through Databox instead.
One tool, limited to MCP data
- Only sees what this MCP server exposes, nothing more
- No connection to your CRM, chat, or revenue data
- Not wired into your business context, so nothing runs automatically
With Databox
Enriched data, the whole story
- The MCP connector brings in this data, same as anywhere else
- Plus your other Databox integrations, already connected
- Your AI Analyst reasons over the whole story, not just this one tool
- Runs automatically with Routines, reusable in every Skill
Related
More like this
Ahrefs
SEO
See the backlink and site audit detail behind an SEO metric, straight from Ahrefs.
Airtable
Productivity
See the record and base behind an operations metric, straight from Airtable.
Asana
Project Management
Bring the task and comment history behind your project metrics into your AI Analyst, straight from Asana.
FAQ
Frequently asked questions
Is there an official GitHub MCP server?
Yes. GitHub runs its own remote MCP server, open to every GitHub account regardless of plan.
Do I need a special GitHub plan to connect it?
No. The GitHub MCP server works on any GitHub account. What your AI Analyst can see still follows your GitHub permissions.
What can my AI Analyst read once connected?
Issues, pull requests, repositories, GitHub Actions runs, code scanning alerts, discussions, and more, scoped to whatever you authorize.
Does Databox have a one-click GitHub connector?
Not yet. Today, connect GitHub as a Custom connector using its official MCP server. A native, one-click Databox connector is planned.
What is an MCP connector?
An MCP connector links your AI Analyst to an external tool through the Model Context Protocol, an open standard for connecting AI assistants to real business systems. Once connected, your AI Analyst can read (and soon act on) the records behind your metrics - deals, tickets, conversations, campaigns - without you switching tabs.
How is an MCP connector different from a Databox integration?
A Databox integration brings your metrics into Databox so you can track, visualize and report on them. An MCP connector gives your AI Analyst access to the records behind those metrics - the deal, the ticket, the ad, the Slack thread - so it can explain why a number moved and act on it in that tool. Many tools, like HubSpot and Facebook Ads, have both, and they work best together: the integration shows what changed, the connector shows why.
Can I connect any tool, or only the ones in the connector library?
Any tool with an MCP server. You can pick an official connector built by Databox, then click connect and sign in. Or you can add a custom connector: paste the server URL, choose how it signs in (OAuth, API key, bearer token or none), name it and connect.
Is my data safe when I connect a custom MCP server?
You decide what each connector can do. Every tool on a connector has its own setting - Always allow, Needs approval or Blocked - and every tool starts on Needs approval. Official connectors are reviewed by Databox. Custom servers are not, and once connected, a server can access your AI Analyst's content and share data outside Databox. Only connect servers you trust. You can change permissions or disconnect at any time.
Do MCP connectors cost extra?
No. MCP connectors are included with your AI Analyst at no extra cost.
What can my AI Analyst do with a connected tool today?
It can read and act. Read tools let it look up deals, tickets, campaigns and messages to explain why a metric moved, answer a question or run a Routine. Write tools let it take action in the tool, like updating a deal stage, sending a Slack message or creating a task. Every action follows the permissions you set, and new tools start on Needs approval.
Give your AI Analyst the full picture
Connect the tools where your business happens, official or custom, and let your AI Analyst explain what changed and why.