ArchForge

Why is your use case a strong fit for WebMCP?

Architecture design is iterative and structured. WebMCP lets an agent safely create, change, validate, and export the same living architecture model a person sees, instead of returning disconnected chat output.

How does it create a better user experience?

People and agents work on one inspectable architecture. The human provides context and makes final choices; the agent makes structured changes and analysis. Recommendations explain rationale and tradeoffs, while Signals shows every agent action and result.

What can people and agents do together that was difficult or impossible before?

Together, they can turn requirements and constraints into a provider-neutral architecture, evaluate AWS or Azure options, identify risks, and export a validated design—all while keeping decisions visible, editable, and reversible.

How did you implement WebMCP?

ArchForge detects document.modelContext and registers typed, runtime-validated WebMCP tools over its application services. UI and agent actions share the same command path; tools return structured results, mutation details, and safe errors. The design_system workflow validates requests before making changes and reports each step.

Built With

  • next
  • vercel
  • webmcp
Share this project:

Updates

Submission history