Hevo MCP Server
# Hevo MCP Server
An MCP (Model Context Protocol) server that exposes Hevo Data's pipeline API as tools for Claude Code and other MCP clients.
## What it does
Manage Hevo Data pipelines directly from Claude Code:
- `list_pipelines` — list all pipelines in your Hevo account
- `get_pipeline` — get details of a specific pipeline
- `create_pipeline` — create a new pipeline
- `enable_pipeline` — enable a pipeline
- `disable_pipeline` — disable a pipeline (optionally cancel active jobs)
- `resync_pipeline` — trigger a resync (optionally drop and reload)
- `update_pipeline_config` — update a pipeline's configuration
- `update_source_config` — update a pipeline's source configuration
## Status
`enable_pipeline`, `disable_pipeline`, and `resync_pipeline` are verified against Hevo's live API docs. `update_pipeline_config` and `update_source_config` have not yet been verified against live docs and may need adjustment. Test carefully before relying on these two in production.
## Setup
### 1. Get your Hevo API credentials
You'll need an API key and secret from your Hevo account, plus your region (`us`, `eu`, `in`, or `asia`).
### 2. Install and build
```bash
git clone https://github.com/felixjoy89hevo/hevo-mcp.git
cd hevo-mcp
npm install
npm run build
```
### 3. Register with Claude Code
```bash
claude mcp add hevo --scope user -- node /absolute/path/to/hevo-mcp/dist/index.js
```
Then set your credentials as environment variables in Claude Code's MCP config (see `.env.example` for the required variables):
- `HEVO_API_KEY`
- `HEVO_API_SECRET`
- `HEVO_REGION`
### 4. Restart Claude Code
After any config or code change, restart your Claude Code session fully. Reconnecting isn't enough.
## Requirements
- Node.js >= 18
## API Reference
Built against [Hevo's live API docs](https://edge-api-docs.hevodata.com/reference/get_api-v1-pipelines).
## License
MIT
TDQS
Scored across 8 tools
Each tool targets a distinct action and object (list, get, create, enable, disable, resync, update pipeline config, update source config). The two update tools are clearly separated by scope (pipeline-level vs source-level) with descriptions that prevent confusion.
All tool names follow a consistent verb_noun pattern in snake_case: list_pipelines, get_pipeline, create_pipeline, enable_pipeline, disable_pipeline, resync_pipeline, update_pipeline_config, update_source_config. The naming style and verb choice are uniform.
With 8 tools, the server covers the essential pipeline management operations without bloat. Each tool serves a distinct, necessary purpose for managing Hevo pipelines.
The surface covers list, get, create, enable/disable, resync, and update operations, which are core to pipeline management. The only notable gap is the absence of a delete_pipeline tool, which is a minor lifecycle gap that agents can work around by disabling pipelines.