OPEN · ALWAYS-ON · BUILT ON FLUE

An autonomous
dev assistant.

Your PR queue, deployments, and review requests, in one always-on console that Neon manages for you. Run it on a companion display, a sensor panel like a Xeneon Edge, or let other agents drive it.

$ npm i -g neondeck@next
★ Star on GitHub Read the docs →

The deck, live.

Sample 32:9 layout for the Corsair Xeneon Edge. Neon can configure itself to run as a companion on any display.

neondeck v0.4.1 │ CPU 12% │ MEM 8.4/32G │ GPU 41% │ 54°C │ NET ↓1.2 ↑0.3 │ flue:online 14:32:07
● GITHUB · @you4 OPEN PRs

flue/harness#1284 · 2h

fix: retry durable steps on cold start ● checks pass◆ 2 review

neondeck/core#312 · 5h

feat: configurable statusline sensors ● 1 failing△ draft

neondeck/panels#88 · 1d

add github review-requests panel ● checks pass↻ rebase
● FLUE AGENT · triage.tsdurable · ctx 18k
What needs my attention this morning?
tool github.list_prs() · 3 results
3 of your 4 open PRs are green. neondeck/core #312 has 1 failing check. flue/harness #1284 has 2 reviewers waiting on you.
fix it
›Message neondeck...Enter send · Ctrl K tools

Built for work in progress.

Neon watches your PRs, prepares fixes when feedback comes in, and keeps them moving.

Your PRs, with CI status at a glance

Every open PR you've authored across your repos in one panel, with live check status and a flag when one goes stale or has issues.

Keep one agent owner with every watched PR

Neon tracks a PR through feedback, checks, and merge with one continuing owner and one managed worktree. It can notify, prepare a commit, wait for approval, or autonomously deliver a validated fix.

Agent + Human reviews = less slop

Review PRs faster, without compromising on thoroughness. Neon performs a pre-review and then hands you short PR briefings, pointers to the parts that matter, and flags issues it noticed. Chat with the agent reviewer as you review the PR.

Super fast diff and review UI

We've got a diff viewer based on @pierre/diffs that can handle large diffs. Leave inline comments, resolve threads, and submit an approval or change request — without switching to github.com.

Briefings and scheduled tasks

A morning briefing pulls your PRs, watches, CI failures, and next actions. Or add a custom instruction and extend the task by pulling in MCPs.

Gated local execution

Repository automation uses managed worktrees, while direct host commands clear an inspectable approval policy first. Coding and validation currently run on your local machine.

Memory that learns from your work

Neon applies typed, audited, reversible memory and skill improvements from conversations and PR outcomes automatically by default. It learns as it watches your review.

Ask Neon to set up the deck

Neon configures itself — repos, models, schedules, even its own layout and display. Want to change something? Just ask.

CONFIG OR CHAT

The layout is a file. Or a request.

Manually tune layouts, text scale, fonts, panel regions, and plugin config in runtime-home dashboard.json, or ask Neon to adapt the deck for your size, orientation, and constraints.

Get started 
{
  "schemaVersion": 1,
  "display": { "width": 2560, "height": 720 },
  "theme": "dark",
  "statusline": { "position": "top", "pluginId": "host-metrics" },
  "layout": {
    "columns": 12,
    "rows": 5,
    "regions": [
      {
        "id": "work", "title": "WORK",
        "column": 1, "row": 1, "columnSpan": 4, "rowSpan": 5,
        "defaultTab": "github",
        "tabs": [
          { "id": "github", "title": "GITHUB", "pluginId": "github-pr-list" }
        ]
      },
      {
        "id": "neon", "title": "NEON",
        "column": 5, "row": 1, "columnSpan": 8, "rowSpan": 5,
        "defaultTab": "chat",
        "tabs": [
          { "id": "chat", "title": "CHAT", "pluginId": "flue-chat" }
        ]
      }
    ]
  }
}