Watching progress

Real-time visibility into what your agent is doing — the Development board and the Activity feed. · Recommended

Once your agent is connected and the first AC is claimed, the Development page is where you live — it's Development in the project sidebar, under Build on an idea-first project and under Ship & Sync on a code-first one. The page carries a two-tab strip: Plan (the sprint plan) and Track (the live board). This page is about Track; for Plan, see Sprint planning.

On the Track tab, two surfaces give you visibility at two different zoom levels. Both update live, with no refresh.

Rolling out. The Stories view, the "now building" line, in-ticket questions and the Releases tab are reaching accounts gradually. If you don't see a Stories | Criteria switch above the board yet, your account is on the classic criteria board described in section 1; everything else on this page still applies.

Figure: The kanban board mid-run.

1. The Development board (mid-zoom)

The board groups acceptance criteria into columns by status:

Cards animate as their status changes, so you see motion as the agent works. If you've planned sprints, the board groups cards into per-sprint sections, with a Foundation section at the top for the scaffolding ACs from your auto-derived Foundation stack.

Stories as tickets

By default the board shows one ticket per user story. Each ticket carries one chip per acceptance criterion, labelled with its code (for example AC-042) and coloured by status, so you can see a story fill in as its tests pass:

A story sits in the column of its most urgent criterion: any blocked criterion moves the whole story to Blocked, any criterion in progress moves it to In Progress, and it reaches Done only when every criterion has been resolved. Click a chip to open that criterion. Use the Stories | Criteria switch to go back to one card per criterion, which is also the view you drag cards in.

Now building

While an agent works on a criterion, its ticket shows a live line: the agent's name, what it is doing, and when it last reported, for example Claude Code · Writing the test · 42s ago. The line turns amber if the agent has been quiet for 3 minutes. After 15 minutes it adds a Nudge button, which copies a ready-made message to your clipboard for you to paste into your agent's session.

Open a criterion to see the rest of what the agent has reported: its plan as a checklist, the files it has touched, how many tests pass, the latest commit with its diff and test-run links, and every earlier attempt with its verdict.

Verified, or accepted by hand

Only a passing CI run marks a criterion Verified (CI). If you pass a criterion yourself it is recorded as Manually accepted, with your note, and it says so everywhere it appears. Criteria that arrived from a code import show Imported (unverified) until a test confirms them.

Questions from your agent

When an agent is unsure about the spec it blocks the criterion and asks. The question appears on the ticket with the agent's suggested answers as buttons, plus a box for your own words. Answering records your decision on the project and unblocks the criterion, and your agent reads the answer on its next check. You also get a notification when a new question arrives.

Approving in bulk

Every criterion starts in Awaiting Approval. Instead of approving them one at a time, use Approve N on a story ticket, or Approve all N awaiting approval at the top of a sprint. Each criterion is still approved individually and recorded as its own event; you see progress as it runs, and if some fail you are told how many and why.

2. The Activity feed (narrative zoom)

Below the board is a collapsible Activity feed — a chronological, human-readable log of every transition:

Agent vm_a1b2c3 claimed User can sign up with email · 2m ago Agent vm_a1b2c3 submitted for review User can sign up with email · 18m ago You approved User can sign up with email · 19m ago

The feed shares the board's live connection, so a new row appears the moment the event is recorded; there is nothing to refresh. If the connection drops, the board shows Polling until it is back and then catches up on everything it missed. When new events arrive while the feed is collapsed, an N new badge appears. Every row names the actor (You, Agent, CI, System), so it doubles as an audit trail, and it shows what the agent said: a progress update reads Claude Code reported progress on AC-042, followed by its summary.

Use the feed to:

• Follow one agent's session end to end.
• Spot blocked work — blocked events stand out in the feed.
• Watch for stalls — an amber now building line, or no new events for 5+ minutes during an active run, usually means your agent has crashed or lost connection.

3. The Handoff panel (setup zoom)

Your connection details are one click away from the board: the ⓘ Handoff Instructions button (on the Track tab, next to the Category filter) opens the Hand off to your IDE dialog. It carries your API-key count with a link to Account → Developer, a one-line curl for the blueprint endpoint, and the MCP config block for your IDE. Reach for it whenever you need to re-connect an agent or check that your key is still live.

Reviewing and approving

When an agent submits an AC it lands in In Review. Resolving it is your job (or your CI's) — agents can't resolve their own work. If your CI reports test results to VibeMap, a passing run resolves the criterion for you and marks it Verified (CI). Otherwise, for each item:

1. Read the diff the agent submitted (a git SHA is required on every submission).
2. Pass → opens a short form. Give a link to a test run, or write a note of at least 10 characters saying what you checked. The form states plainly that this is not CI verification: the criterion moves to Done as Manually accepted, and the next AC in the chain becomes claimable. Dropping a card on Done asks the same thing.
3. Fail → the card is marked failed (it stays in Done); drag it back to Ready and the agent picks it up on its next loop.

💡 Tip: Review in batches every 5–10 minutes, not after every single AC. The agent keeps moving as long as it has claimable work, and you avoid constant context-switching.

Tips for unattended runs

VibeMap's MCP loop is built to run for an hour or more without supervision:

• Check in every 5–10 minutes, approve in batches.
• Use the sprint Gantt totals as a rate gauge — the passed/total count updates live.
• Watch the Backlog. If ACs land in Backlog instead of planned sprints, the planner likely mismatched some slugs — re-plan, or drag them manually.

⚡ Power user: There's no Pause/Resume button because the agent runs in your IDE, not in VibeMap. To pause, tell your agent to stop or close the IDE; to resume, re-issue the "get next ready and implement" instruction — claims persist. For a hard stop, revoke the API key: the next MCP call fails auth and the agent halts cleanly.

↔ The traditional way

Traditionally, "how's it going?" is a standup question, a Slack ping, or a Jira board someone has to remember to update. Here the board is the work: your agent moves its own cards and leaves its own notes as it goes, so progress is something you glance at, not something you chase.

What's next

• Something not moving? → Troubleshooting
• Day-to-day board workflow → Kanban tracker (developer docs)
• Detect spec drift as the code evolves → Drift detection