MCP & AI Tools

Name: Upshot
Brand: Upshot

Upshot exposes your engineering data via the Model Context Protocol (MCP) — so you can ask Claude, Copilot, or any MCP-compatible AI assistant about your metrics.

What is MCP?

MCP is an open protocol that lets AI models call external tools and access real-time data. Think of it as "function calling" but standardized across AI providers.

Available tools

Tool	Description
`get_dora_metrics`	Fetch DORA metrics for a team/repo over a time range
`get_activity_feed`	Recent engineering activity (PRs, deploys, issues)
`search_work_units`	Search across clustered work items by keyword or label
`get_changelog`	Retrieve published changelog entries
`get_contributor_summary`	Activity summary for a specific contributor
`compare_periods`	Compare metrics between two time periods

Connecting via stdio

The fastest way to get started — works with Claude Desktop and other local AI tools:

npx @upshot/mcp

Configure in your AI tool's MCP settings:

{
  "mcpServers": {
    "upshot": {
      "command": "npx",
      "args": ["@upshot/mcp"],
      "env": {
        "UPSHOT_API_KEY": "your-api-key-here"
      }
    }
  }
}

HTTP transport

For server-side or remote connections:

POST https://api-upshot.wellcode.ai/mcp
Authorization: Bearer your-api-key-here
Content-Type: application/json

{
  "method": "tools/call",
  "params": {
    "name": "get_dora_metrics",
    "arguments": {
      "team": "backend",
      "period": "last_30_days"
    }
  }
}

Authentication

Generate an API key from Settings → API Keys in your Upshot dashboard. Keys are scoped to your organization and can be limited to read-only access.

Use cases

"What's our deployment frequency this month compared to last month?"
"Show me what the backend team shipped last week"
"Which contributor has the most open PRs right now?"
"What's our change failure rate trend over the last quarter?"
"Summarize the last 3 changelog entries"