Skip to main content
Glama
hiveflowai

HiveFlow MCP Server

Official
by hiveflowai
README.md
# @hiveflow/mcp-server — deprecated

> **This package is no longer maintained.** Use the hosted Hiveflow MCP server instead:
>
> ## `https://mcp.hiveflow.ai`

The hosted server is always up to date with the platform, signs you in with OAuth (no API key to copy), and needs nothing installed. This repository is archived and the npm package will not receive updates.

*Español: este paquete está deprecado. Usa el MCP remoto `https://mcp.hiveflow.ai`; las instrucciones de abajo sirven igual.*

## Connect your client

**Claude (web, Desktop, mobile)** — Settings → Connectors → *Add custom connector* → URL `https://mcp.hiveflow.ai`.

**Claude Code**

```bash
claude mcp add --transport http hiveflow https://mcp.hiveflow.ai
```

**ChatGPT** — Settings → Apps & Connectors → add a connector with the URL `https://mcp.hiveflow.ai`.

**Cursor** (`~/.cursor/mcp.json`)

```json
{
  "mcpServers": {
    "hiveflow": { "url": "https://mcp.hiveflow.ai" }
  }
}
```

**VS Code** (`.vscode/mcp.json`)

```json
{
  "servers": {
    "hiveflow": { "type": "http", "url": "https://mcp.hiveflow.ai" }
  }
}
```

**Clients that only support stdio** — bridge with [`mcp-remote`](https://www.npmjs.com/package/mcp-remote):

```json
{
  "mcpServers": {
    "hiveflow": {
      "command": "npx",
      "args": ["-y", "mcp-remote", "https://mcp.hiveflow.ai"]
    }
  }
}
```

## Headless use (scripts, CI)

The hosted server also accepts a Hiveflow API key (`hf_...`, created in Settings → API Keys) instead of OAuth:

```bash
npx -y mcp-remote https://mcp.hiveflow.ai --header "Authorization: Bearer ${HIVEFLOW_API_KEY}"
```

## Migrating from this package

| Old tool (this package) | Hosted server |
|---|---|
| `list_flows`, `get_flow`, `create_flow` | same names |
| `execute_flow` | `run_flow` |
| `pause_flow`, `resume_flow` | `set_flow_state` (`paused` / `active`) |
| `list_mcp_servers`, `create_mcp_server`, `get_flow_executions` | removed |
| `blurb_set_emotion`, `blurb_set_state`, `blurb_status` | `hiveflow blurb …` in the [Hiveflow CLI](https://github.com/hiveflowai/hiveflow-cli) (the board is local, a hosted server cannot reach it) |

The hosted server adds organizations, workspaces, kanban boards, cards and workers. Ask your assistant to call the `docs` tool for the full API reference.

## License

MIT

TDQS

B3.4/5.0

Scored across 9 tools

Disambiguation5/5

Each tool has a distinct purpose with clear boundaries: create_flow vs. get_flow vs. list_flows for flow management, create_mcp_server vs. list_mcp_servers for server registration, and execute_flow vs. pause_flow vs. resume_flow for flow execution control. No overlap or ambiguity exists between tools.

Naming Consistency5/5

All tools follow a consistent verb_noun naming pattern (e.g., create_flow, list_flows, pause_flow) using snake_case throughout. The naming is predictable and readable, with no deviations in style or convention.

Tool Count5/5

With 9 tools, the server is well-scoped for managing workflows and MCP servers in HiveFlow. Each tool serves a clear function, and the count is neither too sparse nor bloated, fitting typical expectations for a domain-specific server.

Completeness4/5

The toolset covers core CRUD and lifecycle operations for flows (create, get, list, execute, pause, resume) and MCP servers (create, list), with get_flow_executions providing execution history. A minor gap exists in updating or deleting flows/servers, but agents can work around this for most workflows.

Maintenance

ActivityMaintained
ResponsivenessNo issues