MCP server for lst.so — the task tracker where you and your AI agents share one list.
Manage tasks, subtasks, tags, and activity logs from Claude Code, Cursor, ChatGPT, or any MCP-compatible client.
- Get your API key at lst.so/settings
- Add to your MCP config:
{
"mcpServers": {
"lst": {
"command": "npx",
"args": ["-y", "lst-mcp"],
"env": {
"LST_API_KEY": "your-api-key"
}
}
}
}| Client | Path |
|---|---|
| Claude Code |
.claude/settings.json or ~/.claude/settings.json
|
| Claude Desktop | ~/Library/Application Support/Claude/claude_desktop_config.json |
| Cursor | Cursor Settings > MCP |
-
list_tasks — List tasks. Filter by
status(open/done),tag,agent, ordue(today/overdue/upcoming) - get_task — Get a task with its subtasks and activity log
- create_task — Create a task with priority, agent, due date, tags, notes, and recurrence
-
update_task — Update a task. Use
{done: true}to mark complete - delete_task — Delete a task permanently
- create_subtask — Add a subtask to a task
- update_subtask — Update or check off a subtask
- delete_subtask — Delete a subtask
- add_log_entry — Log progress on a task. Log as you work, not just when done.
- list_tags — List all tags
- create_tag — Create a tag with optional color preset
- delete_tag — Delete a tag
Once connected, your agent can manage tasks naturally:
"Create a task to review the PR, tag it 'code-review', assign to claude-code, due today""What's on my plate? Show open tasks due today"
"Mark task 42 as done and log 'Deployed to production'"
| Variable | Required | Description |
|---|---|---|
LST_API_KEY | Yes | Your lst.so API key from Settings > Agents |
LST_API_URL | No | Custom API URL (defaults to https://lst.so/api/v1) |
MIT