TL;DR: Wrike just got a major upgrade! You can now turn any unstructured request (client briefs, meeting notes, emails, or free-form ideas) into fully structured, execution-ready projects in Wrike through your AI assistant (Claude, ChatGPT, Copilot, Gemini, Cursor) via MCP. Your assistant will identify the right Wrike request form or blueprint, pre-fill all the details, ask clarifying questions for anything missing, and build the entire work breakdown structure automatically, all while honoring your company’s established workflows and business logic.
Hi Wrike Community 👋
We're back with another exciting update with Wrike AI-powered Work Creation.
Today, we're solving one of the biggest friction points in project management: Work Intake.
Converting unstructured briefs, email threads, or meeting notes into a Wrike project usually requires filling out request forms or manually creating each task.
AI-powered Work Creation lets you create work through a conversation with your assistant: from simple tasks to complex structures out of a blueprint.
What’s New?
Instead of creating items manually or navigating forms, you can give your AI assistant your raw input directly. The assistant then handles the setup in Wrike:
- Smart Forms Matching: Your assistant scans your existing Wrike request forms and blueprints to find the exact framework designed for your team's process.
- Auto-Populated Context: Relevant details, custom fields, and task parameters from your prompt or source text are pre-filled automatically. In items created out of Blueprints, a top-level description is filled based on both the template inside the Blueprint and the provided info.
- Intelligent Clarifying Questions: Rather than forcing you through every field in a form, your assistant only asks for the specific missing details needed to begin execution. It’s also able to pull additional information from your Wrike environment and any other tools you connect your AI client to.
- Confirmation & WBS Generation: Before anything goes live, you get a full summary to confirm. Once approved, the assistant builds out the full project structure and subitems in Wrike.
- Customizations intact: MCP prioritises creating work that adheres strictly to your organization's custom item types, workflows, and business rules.
💡 How to Use It (Prompts to Try)
You can speak naturally to your AI assistant: no rigid prompt templates required! Here are a few ways to test it out:
1. From a Client Brief:
"Here is the product launch brief from our client [paste brief]. Use our standard Campaign Request Form in Wrike to set up this project, assigned to the Marketing team lead with target launch dates."
2. From Brainstorming:
"Turn these ideas from the whiteboard into a new project under the Q4 Roadmap using our standard Feature Development blueprint."
Learn more: Klaxoon MCP
3. From a Quick Idea:
"I want to suggest a process improvement and need PMO support for my idea. How do I approach it?"
🔒 Setup & Compatibility
- Works across all MCP clients: Connect directly using MCP with Claude, ChatGPT, Microsoft Copilot, Gemini Enterprise, Cursor, or any custom MCP client while marketplace listings are being finalized. See the MCP documentation for setup instructions.
- No extra setup needed: Work Intake capabilities leverage standard API instructions through your connected MCP server. Your chosen LLM handles the reasoning while remaining strictly tied to your Wrike workspace rules.
💬 We’d love your feedback!
Try creating your next project or task list through your AI assistant and let us know how it goes in the comments below! What workflows or intake templates are you setting up first? Our team is reading every reply!
