productive-mcp-rb2
Enables interaction with Productive.io for task management, time tracking, budget monitoring, and project overview through natural language.
README
productive-mcp-rb2
An MCP (Model Context Protocol) server for Productive.io, extended with delivery-focused tools for budget burn, resource planning, and org overview.
Based on berwickgeek/productive-mcp — extended by Ruben Langeweg.
Quick Start
1. Get your credentials
- Log in to Productive.io
- Go to Settings → API integrations
- Generate a new token
- Note your API token and Organisation ID
To find your user ID: use the whoami tool after setup, or check your profile URL in Productive.io.
2. Configure your MCP host
Claude Desktop
Add to ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) or %APPDATA%\Claude\claude_desktop_config.json (Windows):
{
"mcpServers": {
"productive": {
"command": "npx",
"args": ["-y", "productive-mcp-rb2"],
"env": {
"PRODUCTIVE_API_TOKEN": "your_api_token",
"PRODUCTIVE_ORG_ID": "your_org_id",
"PRODUCTIVE_USER_ID": "your_user_id"
}
}
}
}
Restart Claude Desktop after saving.
Claude Code
claude mcp add productive -- npx -y productive-mcp-rb2
Then set your environment variables in ~/.zshrc or ~/.bashrc:
export PRODUCTIVE_API_TOKEN="your_api_token"
export PRODUCTIVE_ORG_ID="your_org_id"
export PRODUCTIVE_USER_ID="your_user_id"
3. Try it
Once configured, ask your AI assistant:
"Show me my open tasks"
"Log 2 hours to the Gadero project for today"
"Which projects are over 80% budget burn?"
"Mark task 99 as Done"
Environment Variables
| Variable | Required | Description |
|---|---|---|
PRODUCTIVE_API_TOKEN |
Yes | Your Productive.io API token |
PRODUCTIVE_ORG_ID |
Yes | Your organisation ID |
PRODUCTIVE_USER_ID |
Optional | Your user ID — enables my_tasks, me shorthand in all tools, and the productive://me/* resources |
Installation
Via npx (no install required)
npx productive-mcp-rb2
Global install
npm install -g productive-mcp-rb2
Updating
npx users — no action needed. Restart your MCP host and npx resolves the latest published version.
Global install users:
npm install -g productive-mcp-rb2@latest
Local checkout users:
git pull && npm install && npm run build
Then restart your MCP host.
Automated npm publishing
This repo includes .github/workflows/npm-publish.yml.
- Trigger: every push to
main - Guardrail: skips publishing if the current
package.jsonversion already exists on npm - Auth: requires repository secret
NPM_TOKEN
Typical release flow:
npm version patch
git push
Tool Reference
Tools are listed in tables, grouped by area. Read-only tools are safe to call freely; write/mutate tools are noted in the description.
People & Identity
| Tool | Description |
|---|---|
whoami |
Get the current user context — shows which user ID is configured as "me". |
list_people |
List team members. Optional company_id, project_id, is_active, email, limit. |
get_person |
Get details about a person. Requires person_id. |
Projects & Companies
| Tool | Description |
|---|---|
list_projects |
List projects in your organisation. Optional status, company_id, limit. |
list_companies |
List companies (clients). Optional status, limit. |
Tasks
| Tool | Description |
|---|---|
list_tasks |
List tasks. Optional project_id, assignee_id, status, limit. |
get_task |
Get full details of a task. Requires task_id. |
get_project_tasks |
Get all tasks for a project. Requires project_id. Optional status. |
my_tasks |
Get tasks assigned to the configured user. Requires PRODUCTIVE_USER_ID. |
create_task |
Create a task. Requires title. Optional description, project_id, board_id, task_list_id, assignee_id (use "me"), due_date, status. |
update_task_details |
Update title and/or description. Requires task_id. Optional title, description. |
update_task_assignment |
Set or clear a task's assignee. Requires task_id, assignee_id (use "me" or "null"). |
update_task_status |
Set task status by workflow_status_id OR status_name (case-insensitive, partial match). Requires task_id. |
update_task_sprint |
Move task to one or more sprints (custom field). Requires task_id. |
reposition_task |
Reposition a task within its list. Requires taskId. Optional move_before_id, move_after_id. |
move_task_to_list |
Move a task to a different task list. Requires task_id, task_list_id. |
add_to_backlog |
Add a task to the project backlog. Creates a Backlog list if missing. Requires task_id. |
create_tasks_batch |
Create multiple tasks at once. Requires tasks array. |
delete_task |
Permanently delete a task. Destructive. Requires task_id. |
Subtasks
| Tool | Description |
|---|---|
list_subtasks |
List child tasks of a parent. Requires parent_task_id. Optional limit. |
create_subtask |
Create a subtask under a parent. Requires parent_task_id, title. Optional project_id, task_list_id, assignee_id (use "me"), due_date, description. |
Task Todos / Checklists
| Tool | Description |
|---|---|
list_todos |
List checklist items on a task. Requires task_id. |
get_todo |
Get a single todo. Requires todo_id. |
create_todo |
Add a checklist item. Requires task_id, title. |
update_todo |
Rename a todo or toggle completed. Requires todo_id. Optional title, completed. |
delete_todo |
Delete a checklist item. Destructive. Requires todo_id. |
Task Dependencies
| Tool | Description |
|---|---|
list_task_dependencies |
List dependencies for a task. Requires task_id. |
get_task_dependency |
Get a single dependency by ID. Requires dependency_id. |
add_task_dependency |
Link two tasks. Requires task_id, depends_on_task_id. Optional type (blocking |
remove_task_dependency |
Remove a dependency by its ID. Destructive. Requires dependency_id. |
Comments
| Tool | Description |
|---|---|
list_comments |
List comments on a task. Requires task_id. Optional project_id, limit. |
get_comment |
Get a comment by ID. Requires comment_id. |
add_task_comment |
Add a comment to a task. Requires task_id, comment. |
update_comment |
Edit a comment's body. Requires comment_id, body. |
pin_comment |
Pin a comment to the top. Requires comment_id. |
unpin_comment |
Unpin a comment. Requires comment_id. |
add_comment_reaction |
Add a reaction (e.g. like). Requires comment_id, reaction. |
delete_comment |
Delete a comment. Destructive. Requires comment_id. |
Attachments
| Tool | Description |
|---|---|
list_attachments |
List file attachments on a task or comment. Requires task_id or comment_id. |
Boards & Task Lists
| Tool | Description |
|---|---|
list_boards |
List boards in a project. |
create_board |
Create a board. Requires project_id, name. |
list_task_lists |
List task lists. Optional board_id, limit. |
get_task_list |
Get a single task list. Requires task_list_id. |
create_task_list |
Create a task list. Requires board_id, project_id, name. |
update_task_list |
Rename a task list. Requires task_list_id, name. |
archive_task_list |
Archive a task list (reversible). Requires task_list_id. |
restore_task_list |
Restore an archived task list. Requires task_list_id. |
reposition_task_list |
Reposition a task list. Requires task_list_id. Optional move_before_id. |
Folders
Folders group boards within a project.
| Tool | Description |
|---|---|
list_folders |
List folders. Optional project_id, status (1=active, 2=archived), limit. |
get_folder |
Get a single folder. Requires folder_id. |
create_folder |
Create a folder. Requires project_id, name. |
update_folder |
Rename a folder. Requires folder_id, name. |
archive_folder |
Archive a folder (reversible). Requires folder_id. |
restore_folder |
Restore an archived folder. Requires folder_id. |
Workflow Statuses
| Tool | Description |
|---|---|
list_workflow_statuses |
List workflow statuses. Optional workflow_id, category_id (1=Not Started, 2=Started, 3=Closed), limit. |
Time Entries
Time entry creation follows a 5-step workflow: project → deal → service → task → create.
| Tool | Description |
|---|---|
list_time_entries |
View existing time entries. Optional date, after, before, person_id (use "me"), project_id, limit. |
create_time_entry |
Create a time entry (step 5 of the workflow). Requires service_id, time, date, note (≥10 chars). |
update_time_entry |
Update an existing time entry. Requires time_entry_id. Optional time, note, date. |
delete_time_entry |
Delete a time entry. Destructive. Requires time_entry_id. |
list_services |
List all services in the org. |
list_project_deals |
List deals/budgets for a project (step 2). Requires project_id. |
list_deal_services |
List services for a deal (step 3). Requires deal_id. |
get_project_services |
Convenience: get services for a project across all deals. Requires project_id. |
Memberships
| Tool | Description |
|---|---|
list_memberships |
List project/team memberships. Optional project_id, person_id, limit. |
Expenses
| Tool | Description |
|---|---|
list_expenses |
List expenses. Optional person_id (use "me"), project_id, after, before, limit. |
create_expense |
Log an expense. Requires date, amount. Optional project_id, deal_id, note, billable. |
Invoices
| Tool | Description |
|---|---|
list_invoices |
List invoices. Optional company_id, project_id, status, after, before, limit. |
get_invoice |
Get a single invoice. Requires invoice_id. |
Activity & Pages
| Tool | Description |
|---|---|
list_activities |
List activities (changes/updates). Optional project_id, person_id, date range, limit. |
get_recent_updates |
Summary of recent updates with breakdown by item type. Optional days. |
list_pages |
List knowledge base pages. Optional project_id, limit. |
get_page |
Get full content of a page. Requires page_id. |
create_page |
Create a new page. Requires project_id, title. Optional body (HTML), parent_page_id. |
update_page |
Update title and/or body. Requires page_id. Optional title, body. |
move_page |
Move a page under another parent. Requires page_id, target_doc_id. |
copy_page |
Copy a page from a template. Requires template_id. Optional project_id. |
delete_page |
Delete a page. Destructive. Requires page_id. |
rb2 Delivery Tools
These tools are extended beyond the core Productive.io API and are tailored for rb2's delivery workflow.
| Tool | Description |
|---|---|
get_budget_burn |
Budget burn analysis with RAG status (Green <70%, Amber 70-90%, Red >90%). Optional min_burn_pct, project_id. |
get_resource_plan |
Team bookings and utilisation for a date range. Optional person_name, project_id, after, before. |
get_overbooked_people |
Detect team members exceeding capacity. Optional threshold_pct, after, before. |
get_org_overview |
rb2 headcount per subsidiary (NL, SCAPE, Code Blue, CN, PT, NG) with active project counts. |
list_bookings |
Raw resource bookings/capacity entries. Optional person_id (use "me"), project_id, date range. |
Common Workflows
Updating task status by name
update_task_status accepts a status_name instead of an ID — it auto-resolves against the project's workflow statuses (case-insensitive, partial match).
"Move task 12345 to In Progress"
→ update_task_status(task_id="12345", status_name="In Progress")
If the name is ambiguous or missing, the tool returns the available statuses so you can retry with workflow_status_id.
Creating a subtask
"Add a subtask 'Write tests' under task 12345 assigned to me"
→ create_subtask(parent_task_id="12345", title="Write tests", assignee_id="me")
Logging time (5-step flow)
list_projects
→ list_project_deals(project_id)
→ list_deal_services(deal_id)
→ get_project_tasks(project_id) # optional, to link a task
→ create_time_entry(service_id, time, date, note)
The simplest approach is to use the timesheet prompt, which walks through every step automatically.
Managing knowledge base pages
"Create a runbook for project 12345 with the deployment steps"
→ create_page(project_id="12345", title="Deployment Runbook", body="<p>...</p>")
"Move page 999 under page 100"
→ move_page(page_id="999", target_doc_id="100")
Pinning an important comment
list_comments(task_id) → pick the comment ID → pin_comment(comment_id)
Prompts (5 total)
Prompts are pre-built workflow templates that guide the AI through multi-step operations. Invoke them by name in Claude Desktop or with /mcp__productive__<name> in Claude Code.
timesheet
Guided workflow for creating timesheet entries. Walks through project → budget → service → task → time entry selection with proper validation. Accepts optional hints to pre-fill the workflow.
Arguments:
project— project name or ID to start with (optional)date—today,yesterday, orYYYY-MM-DD(optional)time— duration like"2h","120m","1.5h"(optional)work_description— brief description of work performed (optional)
quick_timesheet
Step-by-step guidance for a specific stage of the timesheet workflow. Use when you're already partway through and need help with a specific step: project, budget, service, task, or create.
Arguments:
step— current step:project,budget,service,task, orcreateproject_id— project ID if already selected (optional)deal_id— deal/budget ID if already selected (optional)service_id— service ID if already selected (optional)task_id— task ID if already selected (optional)
weekly_report
Generate a structured weekly status report for a team member: time logged, tasks completed, in-progress work.
Arguments:
person_id— person to report on (leave blank to use the configured user)week_start— start of the week inYYYY-MM-DDformat (defaults to current Monday)
project_health
Full project health check: budget burn, open tasks, team capacity, and recent activity — produces a RAG-rated summary.
Arguments:
project_id— the Productive.io project ID to analyse
sprint_planning
Walk through sprint planning: review backlog, check capacity, select tasks, create new ones, and assign the team.
Arguments:
project_id— the project ID to plan the sprint forsprint_name— name for the new sprint (e.g."Sprint 12")
Resources (6 total)
Resources are read-only data sources you can reference directly. In Claude, prefix with @productive or use the URI scheme productive://.
| Resource | Description |
|---|---|
productive://projects |
All active projects in Productive.io — names, IDs, and status. |
productive://org/overview |
rb2 headcount per subsidiary and total active projects. |
productive://me/tasks |
Open tasks assigned to the configured user. Requires PRODUCTIVE_USER_ID. |
productive://me/today |
Time entries logged today by the configured user. Requires PRODUCTIVE_USER_ID. |
productive://projects/{id}/tasks (template) |
Open tasks for a specific project. Replace {id}. |
productive://tasks/{id} (template) |
Full details of a specific task. Replace {id}. |
Timesheet Workflow
The recommended workflow for logging time follows a strict hierarchy:
list_projects
↓
list_project_deals (get budgets for the project)
↓
list_deal_services (get services for the budget)
↓
list_tasks (optional: find task to link)
↓
create_time_entry (log time with service_id + notes)
The simplest approach is to use the timesheet prompt, which walks through all steps automatically.
Repository
github.com/rubenlangeweg/productive-mcp
License
ISC
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
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.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.