Steep — Claude Code MCP server: what it does, how to install
MCP server linking your Steep workspace to AI agents. It lets you inspect business metrics, review target tracking, and explore team structures in chat.
claude mcp add --transport http mcp https://mcp.steep.app/mcpWhat Steep does
This integration exposes Steep workspace analytics directly to conversational tools. An assistant can retrieve specific metric values across chosen calendar ranges and break down numbers along dimensions. It can also report current progress on company targets or trace entities across existing modules.
Every exposed action is strictly read-only. The integration runs purely as a hosted endpoint at https://mcp.steep.app/mcp rather than a local binary. Authentication relies on OAuth with Dynamic Client Registration, binding assistant permissions to whatever your Steep user account is allowed to see.
What it helps with
- query workspace metrics across custom date ranges and breakdowns
- check progress on active business targets
- look up entities and their linked modules
- list members across workspace teams
Example prompts for the agent
- Show MRR for the past six months segmented by subscription plan
- Check our current progress on the Q2 revenue target
- List the members assigned to the growth team in Steep
How to install Steep
Platform CLI: remote server
claude mcp add --transport http mcp https://mcp.steep.app/mcpPlatform CLI: remote server
codex mcp add steep --url https://mcp.steep.app/mcpConfig file
{
"mcpServers": {
"steep": {
"url": "https://mcp.steep.app/mcp"
}
}
}Config file
{
"servers": {
"steep": {
"type": "http",
"url": "https://mcp.steep.app/mcp"
}
}
}Platform CLI: remote server
gemini mcp add -t http steep https://mcp.steep.app/mcpStep by step, with commands for each platform: How to add an MCP server to Claude Code, Codex, Cursor, VS Code, Gemini CLI and more
Questions about Steep
How to install Steep in Claude Code?
Run claude mcp add --transport http steep https://mcp.steep.app/mcp in your command line. The client connects to the hosted HTTP endpoint and prompts you to sign in via OAuth.
What does Steep MCP server do?
It grants AI models read-only access to your Steep analytics workspace. You can ask for metric trends, inspect target progress, and check team rosters without leaving the terminal.
Does Steep require a local daemon or script?
No local installation is provided. The client connects straight to Steep's managed cloud server at https://mcp.steep.app/mcp over Streamable HTTP.
