Command center
The command center is the page your organization's Owners and Admins open from Reports in the sidebar. It puts the whole organization on one page: every project and its health, what each person and agent is doing, how work is moving through its stages, what is failing or stuck, what the AI costs, what shipped, and how good the results are. Each count opens the Tasks board filtered to the work behind it, so every number can be checked.
It reads your own organization's data only and is on for every plan. Everyone lands on the Tasks board when they open Jaah; Members do not see the command center.
It keeps no data of its own: its counts come from tasks, its rows from projects, people and personas, and its quality figures from Production QA. Every project is Healthy, Watch or At risk, the worst of its factors, worked out afresh on each refresh; a project with nothing to judge shows No data.
Pages in this section
- How the command center works — where its figures come from, how health is decided, and Ask Jarvis.
- Read the command center — the bar, the overview tiles, each tab and how health is worked out.
- Find what needs attention — go from a failed, stuck or at-risk count to the tasks behind it.
- See where your AI spend goes — the Cost tab: spend by project, agent and model, and the spend on failed work.
- Ask the assistant about your organization — ask what is happening, and open the Jarvis view its answer links.
- Check what your team delivered — the Delivery tab: tasks by state, created against done, open epics and what shipped.
- Check your projects' CI and merges — the Releases list, each project's latest CI run and what merged last.
- Move an agent to its best-performing setup — the Best setup panel, and Change default to switch the project's default model.
- Compare how your people use Jaah — the People tab: who starts work, how much lands, what it costs and how fast each person answers the agents.
- Compare agent setups side by side — Compare and Ranking on the Quality tab: two to four setups on the same kind of work, and every setup best first.
- Compare agents and models by outcome — the Performance panel on the Agents tab: each agent, model or configuration's results, and Export CSV.
- Follow up on the lowest-scored work — the AI review on the Quality tab: the score trend, the spread and the tasks it scored lowest, with the reviewer's reason.
- Check whether routing beats the default — the routing results on the Agents tab: how often routing picked the runner, and how its work did against the default.
- Check each project's health — Project health on the Projects tab: each project's last merge, open and done work, and what makes it Watch or At risk.
- See what each agent is doing — the Agents list on the Agents tab: each agent's work in flight, done, success and spend, and the tasks behind them.
- Review anomalies and the weekly digest — the Risks tab's Anomalies and Weekly digest: which figure left its usual level, and where to follow it up.