Dashboard tour

Eight screens. What each one shows, when you would open it, and what the keyboard shortcuts are.

Agents

The list of agents registered against your org. Green dot for healthy, grey for stale (no heartbeat in nine minutes or more), red for revoked. Columns: name, last seen, agent version, os, source count.

Open this first after setup to confirm the first heartbeat arrived. Open it when diagnosing a suspected connectivity issue; see Verifying the connection.

Add machine button generates an auth key. The code displays once, valid for fifteen minutes. Copy it and paste into the agent's setup command; see Auth key setup.

Sources

Every source the agents in your org declared. One row per source, sorted by most recent activity.

Clicking a row opens the source detail: backup and restore-test schedule, retention window, recent backups, recent receipts, drift indicators, the source's timeline (backup and test events across time).

The Run backup now and Run restore test now buttons issue commands the agent picks up on its next heartbeat. Useful for ad-hoc validation after a schema change or migration.

Receipts

Every signed receipt your agents have submitted, newest first. Columns: issued time, source, result (pass / fail / error), agent.

Clicking a receipt opens the detail: the parsed payload, the raw DSSE envelope, per-check results, the summary rows each check returned, and a copy-as-JSON button for forwarding to an auditor who wants to run restorable-verify against it.

Backups

One row per ciphertext blob the agent uploaded. Size, SHA-256, encrypted-with key hash, expiry, source. Backups have a state (active, expired, deleted); expired ones keep their metadata forever for audit purposes and are no longer a candidate for restore tests.

Most customers never open this screen. Open it when you want to confirm a specific ciphertext is in the bucket, or when auditing the retention behaviour.

Events

A chronological feed of everything that happened in your org, typed: agent_registered, backup_created, receipt_issued, backup_failed, restore_failed, access_log, etc. Filter by type or source.

access_log events are the ones to pay attention to longest. Any time a Restorable operator views your org's metadata, an event lands here with the operator id, the action they took, and a reason code (e.g. support_ticket_4321). Your weekly evidence email summarises these; this screen has the full record.

API logs

Every API call against your org's data, from agents and from dashboard users, with status code, duration, actor, and target entity. Useful when debugging "why did the dashboard say X but the API return Y."

Sensitive: API logs show actor ids and endpoints, not request bodies. A compromised admin account does not expose stored requests.

Commands

Commands the orchestrator queued for agents to pick up. One row per command with its type, target agent, issue time, delivery time, ack time, and outcome. Most commands settle in under sixty seconds; longer times indicate a slow or unreachable agent.

Users

The list of users who can access this org's dashboard. Role, last-active. Owners can invite and remove users. Self-serve signup created exactly one user (you); invites let you add more.

Keyboard shortcuts

cmd+k opens the command palette: fuzzy-search sources, agents, receipts. Most daily navigation is faster through the palette than the sidebar.

? shows the full shortcut list. A non-exhaustive sample: g s for Sources, g a for Agents, g r for Receipts, j and k to move through a list, Enter to open, Esc to back out.

Dark theme, by default

The dashboard is dark-theme-only for now. A future preference toggle is on the roadmap, not prioritised.