Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add commands/vmobifystudio/app-dev-team/app-control-roomgit clone --depth 1 https://github.com/vmobifystudio/app-dev-teamWhat it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00024 | $0.00919 |
| Opus 5 | $0.00012 | $0.00460 |
| Sonnet 5 | $0.00005 | $0.00184 |
| Haiku 4.5 | $0.00002 | $0.00092 |
Grade A, and why
app-control-room scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 76 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/app-control-room — The five screens
The product interface. A React app in control-room/, with its own dependencies, served by a
Node-stdlib data plane. It is a projection: nothing on it writes state.
For the zero-dependency diagnostic dashboard — the one that works when this build is broken — use
/app-dashboard. Both exist permanently and neither replaces the other.
Steps
-
Build it once (only needed after a fresh clone or a UI change):
cd "${CLAUDE_PLUGIN_ROOT}/control-room" && npm install && npm run buildIf this fails, say so and point the user at
/app-dashboard. The plugin is correct with the UI absent; a broken frontend build is never a reason a project cannot be inspected. -
Start it from the project root:
node "${CLAUDE_PLUGIN_ROOT}/control-room/server.mjs" --project . --port 4174Print the URL it emits and stop. Do not summarise the screens — the page is the output, and a summary of a live page is stale the moment it is written.
Flags:
--port Nif 4174 is taken ·--no-actionsfor a strictly read-only session.With
dist/absent the server still runs:/stateis live and the page it returns names the two commands that build the UI. It never serves a blank page.
The five screens
| Screen | The question it answers |
|---|---|
| Mission Control | phase · why work is not moving, first · active agents · budget position · latest verification · release readiness. There is no burn-down: on a blocked sprint a burn-down is a flat line that explains nothing |
| Communications | channels and per-ticket threads as a readable conversation, with the structured metadata beside each message (ticket, requirement, decision, artifact updated, verification required). Open questions loudest, flagged when the ticket already shipped. Answers that name no artifact are called out — a closed ledger is not delivery |
| Board | kanban, per-owner load, NEEDS ATTENTION with the reason, and qa (static only) carried visibly in the cell a human reads |
| Team | the activated roster: every role, its trigger, what it is working on, and what is off with its reason. A deactivated role is recorded, never silently absent |
| Founder Inbox | decisions required, each with its context, the recommendation quoted from the log, and the action. Where nobody recorded a remedy it says so — this page never invents advice |
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 2d ago First seen · 76 lines · 24 tokens per session scan A be3197da885e
app-control-room is a command published in the GitHub repository vmobifystudio/app-dev-team (4 stars, last pushed 22d ago), licensed MIT. It adds 24 tokens to every session and 919 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other commands, from other repositories
toh-mobile
Turn the current web app into a mobile app — PWA (default) or native store build (Capacitor).
compose-preview
Generate and verify Compose previews for UI components.
android-build
Build Android project with Gradle, fix errors, generate APK/AAB. Invokes android-build-resolver for issues.
figma-apply
피그마 변경을 현재 design/ 브랜치에 반영(커밋·푸시). /figma-start 이후에 사용.
compose-test
Run Compose UI tests with Espresso. Verify critical user flows.
decompile
Decompile an Android APK/XAPK/AAB/DEX/JAR/AAR and analyze its structure.