Reclaim.ai MCP Server
This MCP server lets MCP-capable clients manage Reclaim.ai tasks through Reclaim's API.
List Reclaim.ai tasks with
activeorallfilters, treatingCOMPLETEtasks as active/pending unless archived or cancelled.Retrieve a specific task by ID.
Create tasks with title and optional fields like priority, deadline, notes, status, category, color, and required time chunks.
Update existing task fields (PATCH-style), with guidance for safely appending notes by fetching and combining current text.
Mark tasks complete/incomplete, permanently delete tasks, and clear scheduling exceptions.
Add scheduled time to a task and log completed work minutes.
Start and stop live timers for tasks.
Prioritize tasks by marking them On Deck.
Expose the active-tasks resource at
tasks://active.
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@Reclaim.ai MCP Server list my active tasks"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Reclaim.ai MCP Server 🚀 (UNOFFICIAL)
⚠️ UNOFFICIAL & UNAFFILIATED – This project is not endorsed, sponsored, or supported by Reclaim.ai. It simply uses Reclaim's public API. Use at your own risk and comply with Reclaim's Terms of Service.
Official alternative: Reclaim.ai offers an official remote MCP server at
https://mcp.reclaim.ai. You may prefer the first-party integration over this community server. See Reclaim's setup instructions and #3.
A community‑maintained Model Context Protocol (MCP) server that lets any MCP‑capable client (Claude Desktop, Continue, Cursor, custom scripts, …) interact with the Reclaim.ai API through a set of standard resources & tools.
🧐 Why MCP?
MCP is the "USB‑C" of LLM integrations – one wire that lets every model talk to every tool.
Run this server once and all your MCP‑aware apps instantly gain Reclaim super‑powers.
Related MCP server: TickTick MCP Server
✨ Key Features
Active‑tasks resource (
tasks://active)13 task‑operation tools (list, create, update, complete, timers, …)
🛡 Type‑safe (TypeScript + Zod) & solid error‑handling
📦 Zero‑config stdio transport – perfect for local AI assistants
📚 MCP Capabilities
Tools (Actions)
Tool | Description | Parameters | ✅ Idemp. | ☠️ Destr. |
| List tasks (default = active) |
| ✅ | ❌ |
| Fetch a task |
| ✅ | ❌ |
| Create a new task |
| ❌ | ❌ |
| Update task properties |
| ✅ | ❌ |
| Mark complete |
| ✅ | ❌ |
| Unarchive / mark incomplete |
| ✅ | ❌ |
| Delete permanently |
| ✅ | ✅ |
| Add schedule minutes |
| ❌ | ❌ |
| Start timer |
| ✅ | ❌ |
| Stop timer |
| ✅ | ❌ |
| Log work time |
| ❌ | ❌ |
| Clear scheduling exceptions |
| ✅ | ❌ |
| Prioritise in planner |
| ✅ | ❌ |
⚠️ Known Issues
COMPLETE ≠ done. Reclaim marks a task COMPLETE when its scheduled block ends, even if you haven't finished the work. This server does include those tasks as active when the LLM uses the tool to pull active tasks (and reminds the model that COMPLETE tasks are still active). However, LLMs (Claude) sometimes ignore COMPLETE tasks when asked for "open" or "active" tasks. If that happens, you may need to prompt the LLM explicitly to "include tasks with status COMPLETE".
🚀 Quick Start
Prerequisites
Node.js
^22.12,^24, or>=26. CI covers the supported Node.js 22 and 24 LTS lines.
Claude Desktop configuration (minimal)
{ "mcpServers": { "reclaim": { "command": "npx", "args": ["reclaim-mcp-server"], "env": { "RECLAIM_API_KEY": "xxx" } } } }
Alternative Configuration:
{
"mcpServers": {
"reclaim": {
"command": "absolute/path/to/node (run `which node` in terminal)",
"args": ["/absolute/path/to/reclaim-mcp-server/dist/index.js"],
"env": { "RECLAIM_API_KEY": "xxx" }
}
}
}Alternative: Manual Installation
If you prefer to install from source:
git clone https://github.com/jj3ny/reclaim-mcp-server.git
cd reclaim-mcp-server
pnpm install && pnpm build
# Run with your API key
RECLAIM_API_KEY=your_api_key node dist/index.js🤝 Contributing
Development uses the Node.js version pinned in .node-version. Runtime and the Vitest 5 harness require Node.js ^22.12, ^24, or >=26.
Bug reports & PRs welcome! To contribute:
Fork the repository
Create a feature branch (
git checkout -b feature/amazing-feature)Make your changes (following the code style)
Commit using Conventional Commits (
feat:,fix:, etc.)Push to your branch (
git push origin feature/amazing-feature)Open a Pull Request
Please squash your commits before opening a PR.
📄 License
MIT – see LICENSE.
Available Tools
13 toolsreclaim_add_timeA
Adds scheduled time (in minutes) to a specific Reclaim.ai task. This blocks more time on the user's calendar. Use this if a task needs more time than allocated (e.g., timeChunksRemaining is 0 but work remains) or if a task has status 'COMPLETE' but the user indicates it's not finished.
Inputs: taskId (required), minutes (required).
reclaim_clear_exceptionsAIdempotent
Clears any scheduling exceptions (e.g., manual adjustments, declines) for a specific Reclaim.ai task, allowing it to reschedule normally.
Inputs: taskId (required).
reclaim_create_taskA
Create a new task in Reclaim.ai. Requires at least a 'title'. Other fields like 'timeChunksRequired', 'priority', 'deadline', 'notes', 'eventCategory' are optional but recommended.
Inputs: title (required), notes, onDeck, status, +7 more.
reclaim_delete_taskADestructiveIdempotent
Permanently delete a specific Reclaim.ai task. This action cannot be undone easily.
Inputs: taskId (required).
reclaim_get_taskARead-onlyIdempotent
Retrieves details for a specific Reclaim.ai task by its ID. Note on 'status': If 'COMPLETE', the scheduled time block ended, but the user has NOT marked the task done. It is still considered active/pending.
Inputs: taskId (required).
reclaim_list_tasksARead-onlyIdempotent
Lists Reclaim.ai tasks. Default filter is 'active'. IMPORTANT NOTE ON 'COMPLETE' STATUS: In Reclaim.ai, tasks marked 'COMPLETE' mean their scheduled time block finished, but the user did NOT necessarily finish the work or mark it done. Treat 'COMPLETE' tasks as ACTIVE and PENDING unless they are also ARCHIVED or CANCELLED. If asked for 'active' or 'open' tasks, YOU MUST INCLUDE tasks with status 'COMPLETE'.
Inputs: filter.
reclaim_log_workA
Logs completed work time (in minutes) against a specific Reclaim.ai task. This reduces the remaining time needed and affects future scheduling.
Inputs: taskId (required), minutes (required), end.
reclaim_mark_completeAIdempotent
Marks a specific Reclaim.ai task as completed/done by the user. This usually archives the task.
Inputs: taskId (required).
reclaim_mark_incompleteAIdempotent
Marks a specific Reclaim.ai task as incomplete (e.g., unarchives it, moves it back to the planner).
Inputs: taskId (required).
reclaim_prioritizeAIdempotent
Marks a specific Reclaim.ai task for prioritization ('On Deck'), increasing its likelihood of being scheduled sooner.
Inputs: taskId (required).
reclaim_start_timerAIdempotent
Starts the live timer for a specific Reclaim.ai task. This indicates the user is actively working on it now and helps log time accurately.
Inputs: taskId (required).
reclaim_stop_timerBIdempotent
Stops the live timer for a specific Reclaim.ai task. Time tracked is automatically logged.
Inputs: taskId (required).
reclaim_update_taskA
Update specific fields of an existing Reclaim.ai task using its ID. This performs a PATCH operation – only provided fields are changed. IMPORTANT: Updating fields like 'notes' overwrites the existing content. To append to notes, you MUST first use 'reclaim_get_task' to fetch the current notes, then provide the full combined text (old + new) in the 'notes' field of this update call.
Inputs: taskId (required), notes, title, onDeck, +8 more.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 13 tools
Each tool targets a distinct action on Reclaim.ai tasks: creating, reading, updating, deleting, time management, prioritization, and status changes. No two tools overlap in purpose; even similar verbs like 'add_time' and 'log_work' are clearly differentiated.
All tools follow a consistent 'reclaim_verb_noun' pattern (e.g., reclaim_create_task, reclaim_list_tasks, reclaim_start_timer). The naming is uniform, with clear verbs that describe the action, making it easy to predict tool behavior.
With 13 tools, the server is well-scoped for task management. Each tool covers a necessary operation without being overwhelming. The count is typical for a domain-specific MCP server, providing sufficient functionality without redundancy.
The server covers core CRUD, lifecycle (complete/incomplete), time tracking (add, log, start/stop timer), prioritization, and exception handling. Minor gaps exist, such as no bulk operations or filtering by date range, but the essential workflows are supported.
Maintenance
Related MCP Connectors
Manage Superlist tasks and lists in plain language from any MCP-compatible AI agent.
Local-first task manager: create, edit, and complete tasks, projects, and checklists via MCP.
Read tasks, habits, events and scheduling links; create and update Reclaim tasks and habits.
Official Akiflow MCP for tasks, calendars, scheduling, time slots, and meeting transcripts.
Related MCP Servers
- AlicenseNot gradedqualityDmaintenanceEnables interaction with Todoist through the MCP interface, providing full CRUD operations for tasks and projects including creating, updating, completing, and filtering tasks with natural language commands.12 npmMIT
- FlicenseNot gradedqualityDmaintenanceEnables interaction with TickTick task management system through MCP, allowing users to view, create, update, complete, and delete tasks and projects via natural language.-
- FlicenseNot gradedqualityDmaintenanceEnables interacting with TickTick tasks and projects through natural language via Claude and other MCP clients.-
- FlicenseNot gradedqualityCmaintenanceEnables natural language interaction with Taskwarrior tasks through the MCP. Supports project-scoped task management including adding, searching, modifying, completing, and annotating tasks.-