GoonIntroduction to Tasks: A New AI Runtime for mCP Run
Tasks is a new AI runtime that allows users to execute prompts and tools on the mCP run platform. In this article, we will explore the capabilities of Tasks and how it can be used to create custom triggers and leverage tools on the mCP run platform.
What is mCP Run?
mCP run is a hosted "app store" and registry for servlets, which are secure and portable tools for AI that can be used in any environment, including desktop, mobile, edge, server, and more. mCP run is built on top of Anthropic's open-source Model Context Protocol, which provides a standard for AI agents and apps to interact with each other.
Demo: Creating an Inbound Lead Router with Tasks
In this demo, we will create an inbound lead router using Tasks. The inbound lead router is designed to help us figure out who to assign a lead to when they submit their information through the website. We have pre-written a prompt for this task, which includes detailed information about how to respond when we get a triggered run and how to use the data that it gets from the contact form submission.
Creating an Inbound Lead Router with Tasks
How Tasks Work
Tasks are a way to leverage all of the tools that are available on the mCP run platform and execute them on the platform with various triggers using a prompt that you write. The prompt can be configured to execute tools in order to figure out how to use the contact form information to get the desired outcome.
Using Tasks to Assign Leads
In this example, we are using Tasks to assign leads to different team members based on the size of the company. We have configured the prompt to extract the domain name from the email, crawl the website, and figure out the size of the company. The prompt then sends a message to the inbound leads channel with the assignment.
Triggering Tasks
Tasks can be triggered manually, or they can be triggered automatically through a URL or a schedule. We can also provide a web hook capability, which allows us to integrate Tasks with other services or microservices.
Example Use Case: Assigning Leads to Team Members
In this example, we are using Tasks to assign leads to team members based on the size of the company. We have configured the prompt to extract the domain name from the email, crawl the website, and figure out the size of the company. The prompt then sends a message to the inbound leads channel with the assignment.
Example Use Case: Assigning Leads to Team Members
Conclusion
Tasks is a powerful new AI runtime that allows users to execute prompts and tools on the mCP run platform. With Tasks, users can create custom triggers and leverage tools on the mCP run platform to build new capabilities for their platforms or software. Tasks can be run like cron jobs, or they can be triggered manually or automatically through a URL or schedule. We encourage you to try out Tasks today and see what you can build. Please go to mCP.run, log in, and create some tasks. We are excited to see what you make.