# Rize MCP Server

Model Context Protocol server for [Rize](https://rize.io) — automatic time tracking for professionals and teams.

## What is Rize?

Rize is an automatic time tracking application that helps professionals and teams understand how they spend their time. It tracks activity in the background, categorizes work automatically, and provides insights into productivity, billable hours, and profitability.

## What is this MCP server?

This is a [Model Context Protocol](https://modelcontextprotocol.io) server that enables AI assistants (like Claude) to interact with your Rize account. It provides 42 tools to manage time entries, clients, projects, contracts, profitability, and team members.

## Available Tools (42)

| Category | Tools | Description |
|----------|-------|-------------|
| **User** | `get_current_user` `get_help` | Authenticated user profile, org info, and tool documentation |
| **Time Entries** | `list_my_time_entries` `list_team_time_entries` `get_time_entry` `create_time_entry` `update_time_entry` `delete_time_entry` | Full CRUD for time entries with date range, client, project, task, and label filters |
| **AI Suggestions** | `generate_time_entries` `regenerate_time_entry` `approve_time_entries` `reject_time_entries` `approve_tag_suggestion` | Generate, review, and approve AI-powered time entry suggestions |
| **Time Analysis** | `get_my_time_allocation` `get_team_time_allocation` `list_my_apps_used` `list_my_events` | Hours breakdown by client/project/task/label, app usage, and tracking events |
| **Clients** | `list_clients` `create_client` `update_client` | Manage client records in your workspace |
| **Projects** | `list_projects` `create_project` `update_project` | Manage projects under clients |
| **Tasks** | `list_tasks` `create_task` `update_task` | Manage tasks under projects |
| **Labels** | `list_labels` | View workspace labels for categorization |
| **Team** | `list_team_members` `update_team_member` `invite_team_member` | Team roster, roles, rates, and invitations |
| **Contracts** | `list_contracts` `get_contract` `create_contract` `update_contract` | Billing contracts with clients (hourly, retainer, fixed fee, hybrid) |
| **Profitability** | `get_org_profitability` `get_contract_profitability` `get_profitability_trend` | Revenue, costs, margins, budget burn, and monthly trends |
| **Billing** | `create_expense` `create_revenue_entry` | Log expenses and recognized revenue against contract periods |
| **Notes** | `add_note` `dictate` | Add notes to time entries or dictate time in natural language |

## Skills

Pre-built workflows that combine multiple tools for common tasks. Use these as prompts with your connected Rize MCP server.

| Skill | Description |
|-------|-------------|
| **Weekly Time Report** | Summarize hours by client and project for the week with daily breakdown and gap detection |
| **Profitability Check** | Analyze org-wide and per-contract profitability, flag at-risk contracts and margin trends |
| **Team Utilization Review** | Compare team members' hours against capacity, surface utilization and burnout risk |
| **Invoice Prep** | Gather billable time entries and expenses for a client and period, formatted for invoicing |
| **Productivity Insights** | Analyze app usage and focus patterns to find productivity improvements |
| **Review Time Entries** | Walk through AI-generated time entry suggestions — approve, reject, edit, or regenerate |

See the [skills on GitHub](https://github.com/rize-io/mcp-server/tree/main/skills) for detailed instructions.

## Connection

### OAuth (Claude.ai / Claude Desktop)

Connect via Claude's MCP integration settings using the server URL:

    Server URL: https://mcp.rize.io/mcp

Claude handles OAuth authentication automatically — no API key needed.

### Stdio (CLI / local)

For local development or CLI usage, install and configure with your API token:

    npm install -g rize-mcp-server

See the [GitHub repo](https://github.com/rize-io/mcp-server) for Claude Desktop, Claude Code, and Cursor setup instructions.

## Chat API

The server also provides a conversational AI chat endpoint that uses the same tools.

    POST https://mcp.rize.io/chat/api

Send a JSON body with `message` (required) and `conversation_id` (optional) fields.
Responds with Server-Sent Events (SSE) streaming the AI response.

## Authentication

All API calls are scoped to the authenticated user. The server uses either OAuth tokens (for hosted connections) or personal API tokens (for stdio/CLI usage). All times are returned in ISO 8601 format and interpreted using the user's configured timezone.

---

Built by [Rize](https://rize.io) | [Documentation](https://docs.rize.io) | [GitHub](https://github.com/rize-io/mcp-server)