Step01
Define Your Tools
Identify the distinct applications or features within your product that can be controlled via the command bar (e.g., 'To-Do App', 'Calendar', 'Search').
AdvancedEngineering
How to Build a Smart Command Bar and App Router with Jev
Create an intelligent command bar (Command-K) that not only finds tools but executes actions within them from a single natural language query. Use a layered decision model to route requests to the right app and then to the right function.
Before you start
Step by step
The workflow
Follow the sequence once, then adapt the prompts, checks, and handoffs to your own setup.
6 steps
Step02
Build the First Layer: Tool Router
Configure the top-level Jev decision call. Its job is to take the user's full text query and identify which of your defined tools is the target. It should handle typos and fuzzy matches (e.g., matching 'to-do' to the 'To-Do App').
Step03
Build the Second Layer: Action Router
For each tool, configure a second Jev decision call. This model's scope is limited to the actions available within that specific tool (e.g., createTask, markAsComplete).
Step04
Process the User Query
When a user types a command, first pass it to the Tool Router (Layer 1). Once the correct tool is identified, pass the rest of the command to that tool's specific Action Router (Layer 2).
Example prompt
Go to the to-do app and mark all the pull requests as low priority.
Step05
Identify Action and Parameters
The Action Router (Layer 2) will parse the remaining text to identify the specific function to call (mark as low priority) and any necessary parameters or targets (all the pull requests).
Step06
Execute the Final Command
With the tool, action, and parameters identified, execute the final command within your application's logic.
This layered or stacked approach is highly scalable. You can build abstractions by wrapping small tools with Jev layers, allowing you to create complex behaviors from simple building blocks.
Build your next product with ChatPRD
Turn an idea into a PRD, user stories, and a plan.
Keep building
Related workflows
intermediateDataJev
How to Deduplicate and Clean Large Datasets Using AI Decision Models
Efficiently clean up large databases by identifying and merging duplicate or near-duplicate records. Use a decision model to get confidence scores for potential matches, allowing for automated merging or manual review.
Open
intermediateProductChatGPTNotion
How to Build a Real-Time Incident Command Dashboard with ChatGPT Sites
Create a custom, real-time incident command dashboard using ChatGPT Sites. This tool pulls live, user-specific data from Slack, Notion, and Google Calendar to keep you informed without disrupting your engineering team.
Open
intermediateProductJev
How to Build a Real-Time Voice-Controlled To-Do List App with Jev
Create a fluid, hands-free to-do list that updates instantly as you speak. Use a multi-step classification process to understand commands, match tasks, and execute functions in real-time without needing to pause.
Open
Start shipping
better products.
Join 100,000+ product managers who use ChatPRD to write better docs, align teams faster, and build products users love.
Free to start
No credit card
SOC 2 certified
Enterprise ready