rocketlane-mcp
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., "@rocketlane-mcp Show me all active projects and their health status."
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.
rocketlane-mcp
An MCP server that exposes your Rocketlane CSM data to Claude and other MCP-compatible AI clients. Ask questions about project health, team workload, tasks, and generate weekly reviews — without leaving your AI assistant.
Tools
Tool | Description |
| All active projects with status, CSM owner, and progress % |
| Per-CSM active project count and overdue task count |
| Recent project updates (scope by project or all, configurable lookback window) |
| All tasks across active projects — filter by project or assignee name |
| All active team members with user ID and email |
| Parse meeting notes into structured task drafts for review |
| Create confirmed tasks in Rocketlane (call after |
| Create a new Rocketlane project |
| Generate a weekly CS review doc and push it to Lark Docs (optional — requires Lark credentials) |
Related MCP server: SmartSuite MCP Server
Quickstart
Claude Desktop (recommended)
Add to ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) or %APPDATA%\Claude\claude_desktop_config.json (Windows):
{
"mcpServers": {
"rocketlane-csm": {
"command": "uvx",
"args": ["rocketlane-mcp"],
"env": {
"ROCKETLANE_API_KEY": "your_rocketlane_api_key_here",
"ROCKETLANE_BASE_URL": "https://api.rocketlane.com/api/1.0"
}
}
}
}Restart Claude Desktop. No separate install step needed — uvx fetches and runs the package automatically.
Install via pip
pip install rocketlane-mcpThen run:
ROCKETLANE_API_KEY=your_key rocketlane-mcpOr create a .env file (see .env.example) and run rocketlane-mcp directly.
Configuration
Variable | Required | Description |
| Yes | Your Rocketlane API key (Settings → Integrations → API) |
| Yes |
|
| No | Lark app ID — only needed for |
| No | Lark app secret |
| No | Lark folder token where review docs will be created |
Example prompts
"Show me all active projects and their health status."
"Which CSM has the most overdue tasks this week?"
"What updates have been posted in the last 3 days?"
"Parse these meeting notes and create tasks for the team: ..."
"Push this week's CS review to Lark."
Development
git clone https://github.com/minhlecsm/rocketlane-mcp
cd rocketlane-mcp
pip install -e '.[dev]'
cp .env.example .env # fill in your credentials
pytestLicense
MIT
Available Tools
13 toolscomplete_taskA
Mark a task as Completed (status value 3). Convenience wrapper over update_task.
For other status changes use update_task with status_value (4=Blocked, 1=To do).
Inputs: task_id (required). Returns: object
create_projectC
Create a new Rocketlane project.
Inputs: project_name (required), owner_id (required), customer_id (required), start_date, +4 more. Returns: object
create_tasksA
Create confirmed tasks in Rocketlane. Call this ONLY after user confirmation.
Each task dict must have:
project_id (str): Rocketlane project ID
task_name (str): Short task subject
description (str): Task details
assignee_user_id (int): userId from fetch_project_members() — never guess
assignee_email (str): emailId of the assignee
assignee_first_name (str): Assignee first name
assignee_last_name (str): Assignee last name
due_date (str | null): YYYY-MM-DD or null
phase_id (int): phaseId from fetch_phases() for this project — never guess
phase_name (str): phaseName matching the phase_id
Returns: List of created task results. Each result includes phase and assignees fields from the API response confirming they were correctly set.
Inputs: tasks (required). Returns: result
delete_taskA
Permanently delete a task. This is a HARD DELETE and cannot be undone.
Call this ONLY after the user has explicitly confirmed they want the task deleted. When in doubt, ask the user to confirm first.
Inputs: task_id (required). Returns: object
get_all_tasksA
Return all tasks across all active projects, enriched with project and customer info.
Inputs: project_id, include_completed, assignee_name. Returns: result
get_csm_workloadA
Return per-CSM active project count and overdue task count.
Inputs: no named parameters. Returns: result
get_projects_snapshotA
Return all active Rocketlane projects with status, CSM, and progress.
Inputs: no named parameters. Returns: result
get_project_updatesA
Return recent project updates with body content.
Each entry has the update title (update_text) plus the full body as
both plain text (update_content, HTML stripped) and the original
rich HTML (content_html). Auto-published status posts have empty
content.
Inputs: project_id, days. Returns: result
get_taskA
Return a single task by ID, enriched with project and customer info.
Inputs: task_id (required). Returns: object
get_team_membersA
Return all active Filum team members with their user_id, name, and email.
Use this to look up valid assignees before calling create_tasks.
Inputs: no named parameters. Returns: result
parse_tasks_from_notesA
Parse a meeting/conversation summary into structured task drafts, grouped by team.
Extracts action items for Filum and Customer teams. Returns draft tasks for user review — DO NOT create tasks yet. After calling this tool, present the draft tasks to the user, ask them to confirm or correct assignees and deadlines, then call create_tasks with the confirmed list.
Inputs: conversation_summary (required), project_id (required). Returns: object
push_weekly_review_to_larkA
Generate weekly CS review and push to Lark Docs. Returns doc_url on success.
Inputs: no named parameters. Returns: object
update_taskA
Update an existing task. Only the fields you pass are changed (partial update).
To change the phase, look up phase_id via the phases for this project (parse_tasks_from_notes returns them) — never guess. To reassign, look up assignee_user_id via fetch_project_members / get_team_members — never guess.
Inputs: task_id (required), task_name, description, due_date, +7 more. Returns: object
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 has a clearly distinct purpose: creating projects vs tasks, updating vs deleting, and various getters for different entities. The only potential overlap is complete_task being a wrapper for update_task, but it is explicitly documented as a convenience for a specific status change, so no confusion arises.
All tool names follow a consistent verb_noun pattern with lowercase and underscores: complete_task, create_project, create_tasks, delete_task, get_all_tasks, get_csm_workload, get_projects_snapshot, get_project_updates, get_task, get_team_members, parse_tasks_from_notes, push_weekly_review_to_lark, update_task. The verbs are action-oriented and the nouns indicate the resource.
With 13 tools, the server covers the core workflows of project and task management without being overwhelming or too sparse. The number is well-scoped for the domain, providing essential CRUD operations, reporting, and utility functions.
The tool set covers project creation, task CRUD, snapshots, updates, team members, and workload reports. However, there is no dedicated tool to fetch project members or phases, which are referenced in create_tasks but must be obtained indirectly via other tools or external calls. This minor gap prevents full self-containment.
Related MCP Connectors
Your AI copilot for customer onboarding — projects, KPIs, tasks, and safe writes.
Connects AI assistants to TrackingTime for querying time entries, projects, tasks, customers, and team data using natural language.
Ask questions about your support tickets, chats, agents and AI answer quality, with cited tickets.
Ask your customer portal and CRM anything, in plain English
Related MCP Servers
- AlicenseBqualityDmaintenanceEnables AI assistants to interact with ClickUp's task management API for core project workflows, supporting operations like task creation, updates, search, assignment, and team analytics through natural language.285MIT
- AlicenseNot gradedqualityNot gradedmaintenanceEnables interaction with SmartSuite workspace data through natural language, providing 6 tools for querying, managing records, exploring schemas, and performing operations with human-readable field names across 10+ configured table types including projects, tasks, clients, and schedules.-
- AlicenseBqualityDmaintenanceEnables AI assistants to query and manage PingCode project management data, including work items, releases, bugs, and requirements through natural language interactions.937 npm6MIT

Coddie MCP Serverofficial
AlicenseNot gradedqualityDmaintenanceSecure read-only access to Coddie project planning data, enabling AI assistants to manage and query projects, phases, and steps.11 npmMIT