Package Exports

This package does not declare an exports field, so the exports above have been automatically detected and optimized by JSPM instead. If any package subpath is missing, it is recommended to post an issue to the original package (@krodak/clickup-cli) to support the "exports" field. If that is not possible, create a JSPM override to customize the exports field for this package.

Readme

cu - ClickUp CLI

A ClickUp CLI built for AI agents that also works well for humans. Outputs Markdown when piped (optimized for AI context windows), interactive tables when run in a terminal.

npm install -g @krodak/clickup-cli && cu init

Talk to your agent

Install cu, install the skill file, and your agent works with ClickUp. No API knowledge needed.

"Read task abc123, do the work, then mark it in review and leave a comment with the commit hash."

"What's my standup? What did I finish, what's in progress, what's overdue?"

"Create a subtask under the initiative for the edge case we found."

"Check my sprint and tell me what's behind schedule."

"Update the description with your findings and flag blockers in a comment."

The agent reads the skill file, picks the right cu commands, and handles everything. You don't need to learn the CLI - the agent does.

Setup

You need Node 22+ and a ClickUp personal API token (pk_... from ClickUp Settings > Apps).

OpenCode

1. Install the CLI

npm install -g @krodak/clickup-cli
cu init

2. Install the skill

mkdir -p ~/.config/opencode/skills/clickup
cp $(npm root -g)/@krodak/clickup-cli/skills/clickup-cli/SKILL.md ~/.config/opencode/skills/clickup/SKILL.md

Claude Code

1. Install the CLI

npm install -g @krodak/clickup-cli
cu init

2. Install the skill

mkdir -p ~/.claude/skills/clickup
cp $(npm root -g)/@krodak/clickup-cli/skills/clickup-cli/SKILL.md ~/.claude/skills/clickup/SKILL.md

Then reference it in your CLAUDE.md or project instructions.

Claude Desktop

Claude Desktop doesn't run CLI tools directly. Use it alongside a coding agent (Claude Code, OpenCode) that has cu installed, or run cu commands manually in your terminal.

Install the CLI:

npm install -g @krodak/clickup-cli
cu init

Homebrew

brew tap krodak/tap
brew install clickup-cli
cu init

Then install the skill for your agent (see OpenCode or Claude Code sections above).

Codex / Other agents

1. Install the CLI

npm install -g @krodak/clickup-cli
cu init

2. Add the skill to your agent

Copy skills/clickup-cli/SKILL.md into your agent's system prompt or project instructions. It's a standalone markdown document that teaches the agent all available commands.

Commands

36 commands total. Full reference with examples and flags.

Tasks

Command	Description
`cu tasks`	List my tasks (filter by status, name, type, list, space)
`cu task <id>`	Task details with custom fields and checklists
`cu subtasks <id>`	List subtasks of a task
`cu create`	Create a task or subtask
`cu update <id>`	Update task fields (status, name, priority, assignee, etc.)
`cu delete <id>`	Delete a task (requires confirmation)
`cu assign <id>`	Assign/unassign users
`cu depend <id>`	Add/remove task dependencies
`cu move <id>`	Add/remove task from lists
`cu open <query>`	Open task in browser by ID or name
`cu search <query>`	Search my tasks by name
`cu duplicate <id>`	Copy a task - coming soon
`cu bulk`	Bulk task operations - coming soon

Sprints & Planning

Command	Description
`cu sprint`	My tasks in the active sprint (auto-detected)
`cu sprints`	List all sprints
`cu assigned`	All my tasks grouped by status
`cu summary`	Standup helper: completed, in progress, overdue
`cu overdue`	Tasks past their due date
`cu inbox`	Recently updated tasks

Comments

Command	Description
`cu comment <id>`	Post a comment on a task
`cu comments <id>`	List comments on a task
`cu comment-edit <id>`	Edit an existing comment
`cu activity <id>`	Task details + comments combined
`cu comment-delete <id>`	Delete a comment - coming soon

Custom Fields

Command	Description
`cu field <id>`	Set/remove custom field values (text, number, dropdown, date, checkbox, url, email)
`cu fields <listId>`	List available custom fields for a list - coming soon

Checklists

Command	Description
`cu checklist view <id>`	View all checklists on a task
`cu checklist create <id> <name>`	Create a checklist
`cu checklist delete <id>`	Delete a checklist
`cu checklist add-item <id> <name>`	Add item to a checklist
`cu checklist edit-item <id> <itemId>`	Edit a checklist item (name, resolved, assignee)
`cu checklist delete-item <id> <itemId>`	Delete a checklist item

Command	Description
`cu tag <id>`	Add/remove tags on a task
`cu tags <spaceId>`	List/create/delete space-level tags - coming soon

Workspace

Command	Description
`cu spaces`	List workspace spaces
`cu lists <spaceId>`	Lists in a space (including folder lists)
`cu auth`	Check authentication status
`cu folders <spaceId>`	List folders in a space - coming soon
`cu members`	List workspace members - coming soon

Time Tracking

Command	Description
`cu time start <id>`	Start tracking time on a task
`cu time stop`	Stop the running timer
`cu time status`	Show the currently running timer
`cu time log <id> <duration>`	Log a manual time entry (e.g. "2h", "30m")
`cu time list`	List recent time entries (default: last 7 days)

Setup

Command	Description
`cu init`	First-time setup wizard
`cu config`	Get/set config values
`cu completion <shell>`	Shell completions (bash/zsh/fish)

Output Modes

Context	Default	Override
Terminal (TTY)	Interactive tables + picker	`--json`
Piped (no TTY)	Markdown (optimized for AI)	`--json` or `CU_OUTPUT=json`

Most commands scope to your assigned tasks by default - keeping output small and relevant for agent context windows.

Configuration

Config file

~/.config/cu/config.json (or $XDG_CONFIG_HOME/cu/config.json):

{
  "apiToken": "pk_...",
  "teamId": "12345678"
}

Environment variables

Environment variables override config file values:

Variable	Description
`CU_API_TOKEN`	ClickUp personal API token (`pk_`)
`CU_TEAM_ID`	Workspace (team) ID
`CU_OUTPUT`	Set to `json` to force JSON output when piped (default: markdown)

When both are set, the config file is not required. Useful for CI/CD and containerized agents.

Why a CLI and not MCP?

A CLI + skill file has fewer moving parts. No server process, no protocol layer. The agent already knows how to run shell commands - the skill file teaches it which ones exist. For tool-use with coding agents, CLI + instructions tends to work better than MCP in practice.

Development

npm install
npm test          # unit tests (vitest, tests/unit/)
npm run test:e2e  # e2e tests (tests/e2e/, requires CLICKUP_API_TOKEN in .env.test)
npm run build     # tsup -> dist/

@krodak/clickup-cli

Package Exports

Readme

cu - ClickUp CLI

Talk to your agent

Setup

Commands

Tasks

Sprints & Planning

Comments

Custom Fields

Checklists

Tags

Workspace

Time Tracking

Setup

Output Modes

Configuration

Config file

Environment variables

Why a CLI and not MCP?

Development