∞ · Element Cu

clickup cli

A token-efficient CLI and MCP server for the ClickUp API. ~130 endpoints. 143 MCP tools. ~98% smaller than raw JSON. Works with any AI agent.

v0.6.7 · Rust · Apache-2.0

The Problem

Why This Exists

ClickUp's API returns deeply nested JSON. A single task list query produces ~12,000 tokens of data — statuses wrapped in objects, assignees as arrays of user objects, timestamps as Unix milliseconds, custom fields, checklists, dependencies.

For AI agents operating within context windows, this is a serious problem. A few API calls can consume most of the available context, leaving little room for reasoning.

You didn't build an AI agent to waste 12,000 tokens on a task list. There's an element for that.

The Solution

Token-Efficient Output

clickup-cli flattens nested objects, selects only essential fields, and renders compact output — whether you're using the CLI or the MCP server.

📊

Raw API JSON

~12,000 tokens for 5 tasks. Nested status objects, full user profiles, timestamp integers, custom field metadata.

⚡

clickup-cli Output

~150 tokens for 5 tasks. Flattened status strings, comma-joined assignees, formatted dates. ~98% reduction.

🔧

Same Data, Less Noise

status: "in progress" instead of status: {status: "in progress", color: "#4466ff", ...}

Coverage

28 Resource Groups, 4 Utilities

Core

workspace · space · folder · list · task

Collaboration

comment · checklist · tag · field · task-type · attachment

Tracking

time · goal · view · member · user

Communication

chat (v3) · doc (v3) · webhook · template

Admin

guest · group · role · shared · audit-log · acl

Utilities

setup · auth · status · completions · agent-config · mcp

Quick Start

Get Running

npm

npm install -g @nick.bester/clickup-cli

Homebrew

brew tap nicholasbester/clickup-cli brew install clickup-cli

Cargo

cargo install clickup-cli

Docker

docker build -t clickup-cli .

# Configure
clickup setup --token pk_your_token_here

# Verify
clickup auth whoami

# Use
clickup task list --list 12345
clickup task create --list 12345 --name "My Task" --priority 3

AI Integration

Two Ways to Connect

CLI Mode (recommended)

Most token-efficient. Inject a ~1,000 token command reference into your agent instructions. The agent runs CLI commands directly. Works with any LLM.

clickup agent-config inject

Auto-detects CLAUDE.md, agent.md, .cursorrules, .github/copilot-instructions.md

MCP Mode (143 tools)

For Claude Desktop, Cursor, and tools that prefer native tool integration. 143 tools with compact responses.

clickup agent-config init --mcp

Creates .mcp.json at project root with auto-detected binary path

Installation → Commands → MCP Server →