New request
#721: Visual Reference: Hermes Monitor Dashboard Examples for feature requests: #687 #688 #689 #690 #691 #692 #693 #694 #696 #697 #698 #716 #719
We'll provision a sandbox, run an agent against the issue, and open a draft PR. You can pull the branch and iterate from there.
For visual reference across all of the above feature requests, I have built a working prototype dashboard called Hermes Monitor that demonstrates exactly how all these features should look and behave together. Please use it as a design and layout reference when scoping and building each feature.
The dashboard has the following tabs and components:
Dashboard tab — KPI cards (tasks completed, daily cost, hallucinations detected, tokens used), agent status table with quality scores and hallucination rates per agent, model replacement modal, 7-day cost bar chart with budget tracking and monthly pace projection, model health comparison table, and a live event stream log
Agents tab — full swarm hierarchy showing the Orchestrator at the top, with teams arranged in a 3×5 grid (up to 15 teams), each team showing its named Team Lead and agents beneath with role, title, model assignment, and skill tags. References issue #690 for the tree view and #719 for the teams panel
VPS Health panel — CPU, RAM, and Disk live indicators in the sidebar, referenced in #693
Hallucination alert banners — dismissible warning banners when an agent exceeds the hallucination threshold, referenced in #688
The HTML file for the dashboard is attached to this issue.
Design reference artifact and prototype example attachment supporting related dashboard feature requests.
- docs/design/dashboard-reference.html
- docs/examples/monitor-dashboard.md