Dashboard
Dashboard is the landing page at https://<your-org>.kcap.ai/. It brings together your recent work and the actions that need your attention. Choose a repository or project in the Scope selector to narrow the view.
Read your last seven days
Section titled “Read your last seven days”Your Last 7 Days shows session count, active time, tokens, cached tokens, and estimated cost, with comparisons to the preceding period where available. A generated summary can explain the activity above the numbers.
Active time adds up session durations; concurrent sessions can add up to more than the elapsed wall-clock time. Cost is an estimate from recorded usage and known model prices. A missing value means data is unavailable, rather than zero usage.
Decide what to do next
Section titled “Decide what to do next”What To Do Next links recommendations to the work behind them: a work item, session, or PR review. Some recommendations offer an action such as starting a hosted agent to work on a plan. Read the action before choosing it; launching an agent starts work on a connected daemon.
Needs You surfaces agents waiting for input or permission and work requiring attention. Open the card to answer the request or inspect the failure.
Pick up ongoing work
Section titled “Pick up ongoing work”- Pick Up Where You Left Off brings recent active work back into reach.
- Your Work Items groups the issues, PRs, and sessions you are working on.
- Recent Outcomes summarizes what finished. Open an outcome to read the session and its evidence.
- Recently Learned links to knowledge produced by your team’s work.
- Pending Approvals groups outstanding requests, such as project membership and knowledge decisions.
Sections depend on what is available in the selected scope. An empty queue is not a reason to re-import sessions.
Go deeper
Section titled “Go deeper”| Open | To |
|---|---|
| Insights | Ask questions about activity, cost, tool use, and evaluation trends. |
| Sessions | Find a conversation and inspect its turns, events, and evaluation. |
| Agents | Launch or continue a hosted agent and answer permission requests. |
| Work items | Follow a piece of work across sessions, issues, and PRs. |
| Knowledge | Review findings and maintain reusable guidance. |
| Evals | Inspect quality scores and evaluation coverage. |
| Flows | Follow structured agent runs and review rounds. |
The sidebar’s Search opens the command palette. Use it when you know what you want to find and do not want to browse a list.
Live updates
Section titled “Live updates”Sessions, agents, outcomes, and attention queues update as new data arrives. If the connection drops, the app reconnects and refreshes its state. A catch-up indicator means the server is still processing recorded events; newly captured work may take time to appear.