@isteamhq/mcp
MCP server for is.team, enabling AI agents to interact with project boards, tasks, cards, sprints, integrations, and real-time notifications.
README
@isteamhq/mcp
MCP server for is.team — connect AI agents to your project boards.
Setup
Claude Code / Claude Desktop
Add to your MCP config (.mcp.json or Claude Desktop settings):
{
"mcpServers": {
"is-team": {
"command": "npx",
"args": ["-y", "@isteamhq/mcp"],
"env": {
"IST_API_TOKEN": "ist_your_token_here"
}
}
}
}
Get your API token
- Go to is.team and open Account Settings
- Navigate to the API tab
- Generate a new token
- Copy the
ist_xxxtoken into the config above
Tools
v3.0 ships ~113 tools total — task & card lifecycle (33), canvas (12), workspace primitives (9), team ops (6), sprints (5), niche (12), real-time/chat (5), plus 34 integration tools (GitHub, Drive, Slack, Figma, Calendar). The full surface is grouped below.
Card & task essentials
| Tool | Description |
|---|---|
list_cards |
List cards with LLM access enabled |
read_card |
Read card content — tasks, details, connected cards |
create_task / update_task / complete_task / move_task / delete_task |
Task lifecycle |
reorder_tasks / restore_task |
Re-order tasks; un-archive a task |
add_comment / update_comment / delete_comment / list_comments |
Comment CRUD |
log_time / list_worklogs / get_worklog / update_worklog / delete_worklog / get_active_timer |
Worklog & timer |
add_task_attachment / remove_task_attachment / list_task_attachments |
File attachments |
add_task_subscriber / remove_task_subscriber / list_task_subscribers |
Watchers |
add_subtask / remove_subtask / list_subtasks |
Subtask hierarchy |
list_all_tasks / list_archived_tasks |
Cross-section task listing |
Boards & cards (canvas columns)
| Tool | Description |
|---|---|
list_boards / create_board / update_board / delete_board |
Board CRUD |
create_card / update_card / update_card_settings / delete_card |
Card CRUD + automation rules (autoAssignee, autoLabel, aiAutomationPrompt, …) |
Notes / canvas
| Tool | Description |
|---|---|
list_notes / create_note / update_note / delete_note |
Note CRUD (orphansOnly flag for cleanup) |
create_edge / delete_edge |
Edge CRUD |
create_stack / add_to_stack / dissolve_stack |
Note stacks |
move_node / batch_move_nodes / batch_delete_nodes |
Single + bulk canvas ops |
Workspace / team
| Tool | Description |
|---|---|
list_members / update_member_role / remove_member / update_member_profile |
Member ops |
invite_members / list_pending_invites / revoke_invite |
Invitations |
get_workspace_settings / update_workspace_settings |
Workspace metadata |
get_user_preferences / update_user_preferences |
Caller's prefs (timezone, email, canvas) |
list_files / delete_file |
Workspace file library |
Sprints (agile flow)
| Tool | Description |
|---|---|
create_sprint / update_sprint / complete_sprint |
Sprint lifecycle (with snapshot) |
list_sprints / get_active_sprint |
Read sprints |
Forms (read-only)
| Tool | Description |
|---|---|
list_forms / get_form / list_form_submissions |
Inspect forms + submissions |
Notifications
| Tool | Description |
|---|---|
list_notifications / mark_notification_read / delete_notification |
Caller's notification feed |
Real-time + chat
| Tool | Description |
|---|---|
subscribe_card / unsubscribe_card |
Real-time card watcher |
chat_respond / chat_history / ask_chat |
Card AI chat |
Integrations (34 tools)
GitHub (12), Google Drive (8), Slack (7), Figma (3), Google Calendar (5). See list_integrations.
Real-time notifications
Use subscribe_card to watch a card for new tasks. When a task is created or moved into the card, you'll receive an automatic notification — no polling needed.
"Subscribe to card col-xxx and work on any new tasks that appear."
The AI agent will receive instant notifications and can start working immediately.
Background daemon mode (v2.0+)
Run Claude as a persistent background daemon that auto-executes any task assigned to a chosen card. The daemon survives terminal closure, restarts itself on crashes, and relays Claude's output back to the card chat so you can monitor progress entirely from is.team.
Agent name (v2.1+)
Every setup run now asks for a 6-character agent name (letters + digits). The name appears on the agent's badge in the is.team dashboard, so when you run multiple agents — one per project, one per machine, one foreground and one in background — you can tell them apart at a glance. Use different names for each terminal/project. Examples: HOME01, MACM01, DEV001, LAPTP1, PROD01.
The name is stored in the project's .mcp.json under the IST_AGENT_NAME env var (and in ~/.isteam/daemon.json when running in daemon mode). If you ever need to change it, re-run setup or edit the file by hand.
Setup
npx @isteam/mcp@latest setup --token ist_xxx
When asked Run in background as a daemon?, answer y. The wizard will:
- List your cards — pick the one the daemon should watch
- Ask for a permission mode (
acceptEditsrecommended) - Ask for a working directory (defaults to current)
- Install a launchd agent (macOS) or systemd user unit (Linux)
- Start the daemon
Now assign a task to the chosen card and Claude will start working automatically — no terminal needed.
Managing the daemon
npx @isteam/mcp daemon status # show current state + config summary
npx @isteam/mcp daemon logs --follow # tail the live log
npx @isteam/mcp daemon start|stop|restart
npx @isteam/mcp daemon uninstall # remove the service
How it works
- macOS:
~/Library/LaunchAgents/team.is.mcp-daemon.plist - Linux:
~/.config/systemd/user/isteam-mcp-daemon.service - Config:
~/.isteam/daemon.json(0600 — contains your API token) - Logs:
~/.isteam/daemon.log+~/.isteam/daemon-error.log
Each task spawns a one-shot claude --print subprocess with the repo's MCP config loaded. Claude reads the task, executes it, posts a summary back to the card chat, and moves the task to the next pipeline card.
Platform support
| Platform | Status |
|---|---|
| macOS (launchd) | ✅ supported |
| Linux (systemd) | ✅ supported |
| Windows | ⚠️ not yet — use WSL2 or foreground mode |
Enable LLM access on a card
Before an AI agent can interact with a card:
- Click the AI Integration button on the card header
- Enable LLM Access
- Enable Flow Actions (for create/update/move/complete)
- Enable Comments (for adding comments)
Environment variables
| Variable | Required | Description |
|---|---|---|
IST_API_TOKEN |
Yes | Your is.team API token (ist_xxx) |
IST_AGENT_NAME |
No | 6-char agent badge (A-Z, 0-9). Set by setup wizard. Falls back to a random 6-char id if missing or malformed. |
IST_BASE_URL |
No | API base URL (default: https://is.team) |
License
MIT
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.