Skip to Main Content

Claim Coordinator Agent Demo This project turns the workflow design for a claim-coordination agent into a working local proof-of-concept. It models the claim lifecycle, action-state tracking, evidence-based task completion, and the review gates described in the workflow. What is included • A lightweight Node.js + Express API that stores claims in memory. • A task graph that mirrors the mock claim lifecycle from intake through close. • Action-state tracking so a task is only considered done after the required proof is recorded. • A simulation UI for the main workflow events and human approval gate. • A test suite covering the happy path to a closed claim. Run locally npm install npm start Open http://localhost:3000 to view the dashboard. Useful workflow events The UI exposes a set of demo events that move the claim through the lifecycle: • Claim filed • Insurer opened • Rental covered • Rental booked • Body shop booked • Estimate arrived • Estimate submitted • Estimate approved • Repair authorized • Repair ready • Rental extended • Medical invoice • Medical paid • Settlement offer • Approve settlement • Rental returned • Reconcile and close Validate npm test Workspace packages The repo is also a pnpm workspace (pnpm install at the root): • packages/db (@claim-coordinator/db): the MongoDB data layer. Run pnpm db:setup, pnpm db:seed, pnpm db:smoke. See packages/db/README.md. • apps/dashboard: the web UI on :3000. Claimants start a claim and follow it; operators watch the task graph and action timeline, decide approvals and run the demo panel. Run pnpm --filter dashboard dev next to pnpm agent:dev. See apps/dashboard/README.md. • apps/agent: the Claim Coordinator agent. It has mock insurer/body shop/rental/clinic systems on :4000 and the control API, webhooks and phone line on :4100, plus the supervised workers. Run pnpm agent:dev. See apps/agent/README.md for endpoints and the demo scenes (pnpm --filter agent demo:all). The Express prototype above (npm start) is separate from those and keeps its state in memory. It also uses :3000, so don't run it alongside the dashboard.

Team

More from Agents in Context Toronto: a MongoDB Hackathon

We simulate real world agent to agent systems built on MongoDB, Langchain (harnesses) ready for use in production to defend all systems. A swarm of real AI…

Legacy Siege project preview

suture is an incident-recovery agent that finishes the job even when its worker crashes mid-run. Cleaning up a leaked key takes a chain of writes across…

suture project preview

Venues play one playlist for everyone, and the apps that let guests shape the music keep their likes on view with no easy way to take them back. Musory: your…

Musory project preview

Robots are starting to learn from people, and fleets can share what one robot learns with every other robot. Nobody can yet answer three questions: who taught…

One Lesson project preview

Phylax Tower is a safety net for airport control towers. In March 2026 at LaGuardia, a jet was cleared to land, and two minutes later a fire truck was cleared…

Phylax Tower project preview

New hires hit the ground at light speed. New devs lose their first days juggling broken setup guides, each solving the same problems alone. Day One finishes…

Day One project preview