Dashboard testing

Registered agents appear in the embedded Runiq Dashboard through runtime metadata.

The dashboard is served by the same ASP.NET Core application. It can list agents, show their instructions and attached tools, run chat requests, stream assistant output, and inspect tool call details.

Enable the dashboard

using Runiq.Core;

var app = builder.Build();

app.UseRuniqDashboard(options =>
{
    options.Path = "/dashboard";
    options.Title = "Runiq Workflow Travel Planner";
});

app.Run();

The default path is /runiq. If you set Path = "/dashboard", open /dashboard in the same host application.

What the dashboard reads

Agent visibility comes from runtime metadata and execution events.

Dashboard examples

Registered agents discovered by the Runiq runtime.

Weather Agent in the embedded Runiq Dashboard

Testing Planner Agent

The Workflow Travel Planner sample registers three agents: Weather Agent, Places Agent, and Planner Agent.

For dashboard testing, start with Planner Agent. It is the best user-facing example because it is responsible for the final itinerary and synthesis. Weather Agent and Places Agent are specialist contributors in the workflow.

After registering PlannerAgent.Create(openAiApiKey), open the dashboard and select Planner Agent.

Use an input that gives the planner enough context to produce a final answer:

Create a practical two-day Istanbul trip plan. Keep it low-fatigue. Include weather-aware suggestions, walkable areas, rest breaks, and meal suggestions.

Expected behavior:

You can also use a prompt that looks like the output of previous workflow steps. This tests Planner Agent as the final synthesis step:

Weather context: Istanbul is partly cloudy and comfortable for walking with a light jacket.
Places context: Prefer Sultanahmet, Galata, Karakoy, and Kadikoy. Keep walking distances moderate.

Create the final two-day itinerary with meal suggestions.

Expected behavior:

Testing specialist agents

Weather Agent and Places Agent are still useful to test, but they should read as specialist agents rather than the main product example.

For Weather Agent:

I am planning a light walking day in Istanbul. Give only the weather and comfort contribution for the planner.

Expected behavior:

For Places Agent:

Suggest walkable places for a low-fatigue Istanbul plan. Give only the places contribution for the final planner.

Expected behavior:

Testing the three agents side by side makes the workflow model clearer: Weather Agent and Places Agent create specialist contributions, while Planner Agent owns the final user-facing itinerary.

Troubleshooting