What is ScenarioForge?
ScenarioForge is an interactive personal decision simulator where a human and an AI agent collaboratively build, modify, project, and compare live “what-if” scenarios on the same page.
Why WebMCP is a strong fit
Traditional AI advice stops at text. ScenarioForge gives an agent structured tools that let it work directly with a live decision model. Instead of describing what a user should change and making the user manually update a spreadsheet, an agent can create scenarios, change variables, project outcomes, and compare alternatives through the web app.
Better user experience
The human stays in control of the decision while the agent handles repetitive modeling work. Changes become immediately visible in the interface, so users can inspect assumptions and consequences instead of trusting a black-box recommendation.
What people and agents can do together
A person can provide the real-world constraints and judgment while an agent performs structured scenario operations. Together they can create a baseline, test alternatives, modify inputs, project results, and compare scenarios without repeatedly copying numbers between chat and a spreadsheet.
How WebMCP was implemented
ScenarioForge is a static vanilla HTML/CSS/JavaScript application with a shared client-side scenario store. WebMCP tools are registered through the page's model context and expose structured operations for working with scenarios. The tools operate on the same live state that the human UI uses, allowing human and agent interactions to converge on one model.
Submission-period work
ScenarioForge was developed and hardened specifically for the WebMCP Challenge. The project was refined around agent-accessible scenario operations, live projections and comparisons, and a simple static deployment that judges can inspect directly in the browser.
Log in or sign up for Devpost to join the conversation.