Give every product a workspace
Group its repos, tasks, terminals, browser tools, and saved layouts in a setup that returns exactly as you left it.
CmdBrief pins agents, tasks, terminals, diffs, and browser context to each product—so you always know what is working, blocked, or waiting for you.
One product needs a build fix. Another needs launch copy. A campaign is ready to review while yesterday’s release is producing traffic data. CmdBrief gives each product a home and rolls up the work that needs you, so switching products does not mean rebuilding the operating context.
Group its repos, tasks, terminals, browser tools, and saved layouts in a setup that returns exactly as you left it.
Build with Claude Code or Codex, work through launch tasks, and keep growth or traffic-review tools open beside the product context.
Agent status, attention badges, tasks, and native notifications show which product can move as soon as you act.
Ask what is running, blocked, or waiting; start or resume any supported agent; and dictate the next instruction without navigating between panes.
Inspect agent status, workspaces, usage limits, approved terminal output, and project files.
Use a language supported by the selected provider, choose the assistant’s reply language, and have instructions translated into technical English before they reach the agent.
Workspace reading is session-scoped, and every state-changing action waits for local confirmation.
OpenAI Realtime or xAI Grok · Your API key · Provider usage billed separately · Language availability and accuracy vary by provider.
Agent status, session resume, rate limits, a real terminal, every tool, and git state — all in one local window.
Claude Code, Codex, Grok CLI, Kimi CLI, OpenCode, and Cursor CLI report what they are doing, per pane, per tab, per workspace. Attention rolls up so a blocked agent cannot disappear behind a tab.
Status comes from three layered signals: agent hooks (Claude Code, Codex, Grok, Kimi, and Cursor CLI; authoritative, append-only JSONL), process liveness, and a screen-scrape fallback (how OpenCode is tracked).
Your recent local agent sessions from every supported CLI, listed with the context you need to recognise them, and reopened in one click.
| Session | Runtime |
|---|---|
| Extend the pane position picker to dock panes beside specific panes | 22m |
| Stop forwarding keyboard-activated clicks to mouse-tracking TUIs | 14m |
| Add a hover scrubber with value and time callout to panel charts | 41m |
| Clean a Rust target folder straight from the disk analyzer | 9m |
Reopens with claude --resume or codex resume. CmdBrief reads session metadata plus short, bounded prompt and response previews — never full transcripts.
The usage windows each provider actually reports, side by side, so you can decide where limited agent capacity does the most good.
Windows are whatever each provider actually reports — 5-hour and weekly for Claude Code and Codex, weekly credits for Grok CLI, rolling plus weekly and monthly for OpenCode Go, monthly Cursor-models and Other-models pools for Cursor CLI. Kimi CLI exposes no usage data, so it shows none. Toggle between Remaining and Used; rollover is detected once a reset passes.
Not a chat box with a shell bolted on. A hand-written xterm-class VT emulator on a GPU cell renderer, with every command annotated as it runs.
Every command is annotated with its working directory, duration, exit status, and a git snapshot taken when it ran.
A workspace per product, tabs inside it, and a recursive pane tree you can split, drag, zoom, and restore exactly as you left it.
Panes split recursively with weighted dividers, snap guides, and drag-and-drop rearrangement. Esc restores a zoomed pane.
A per-repo dashboard and a dedicated diff pane, both entirely local. CmdBrief never pushes, fetches, or phones home.
Branch, upstream, ahead/behind, staged, unstaged, untracked and ±lines, plus the last 30 commits with diff previews. No network calls.
Every product, project, and agent stays visible.
Repos and workspace state stay on your Mac; connected services remain explicit.
One persistent context through every product stage.
Turn the next action into a running agent.
Building, launching, marketing, and reviewing results use different tools. CmdBrief keeps the repo, files, local services, browser context, tasks, and agent runs organized around the product they serve.
CmdBrief stands on fast terminal fundamentals: GPU rendering, durable scrollback, split panes, command blocks, and git-aware context.
Save tabs, panes, layout, and startup commands as a workspace you can reopen for any product. Secrets stay out of the file as ${NAME} references.
layout.yamlCmdBrief is itself an MCP server. Your agent can inspect and reconfigure workspaces — every change staged and applied only after you approve it in-app.
--cmdbrief-mcpPer-workspace tasks carry status, priority, and acceptance criteria. Assign the next job to yourself, Claude, or Codex without detaching it from the product.
status · priorityA native macOS menu-bar readout with live CPU, GPU, RAM, and battery. Agent attention surfaces as in-app badges and native notifications.
NSStatusBarImported layouts have auto-run commands demoted to manual, so a shared workspace can never silently execute anything.
auto-run demotedWeighted splits with snap guides, pane zoom, drag-and-drop rearrangement, and layout presets from single pane to 2×2.
2×2 presetsEach command block records the git snapshot it ran against. The dashboard is a local-only view — it never pushes, fetches, or phones home.
git snapshotA Rust core with a WGPU renderer whose glyph pipeline is matched to the UI, plus a native Swift menu-bar layer. Fast under load, easy on the battery.
WGPUThe tools change at each stage. CmdBrief keeps the product, evidence, tasks, and agent work connected from build through review.
Explore founder use casesKeep the repo, local app, implementation tasks, coding agents, and review evidence together from first commit through a working release.
Keep release checks, launch tasks, browser previews, copy changes, and agent follow-ups attached to the product that is going live.
Track campaign and content tasks beside the files, browser tools, and agent sessions doing the work instead of rebuilding context in every app.
Open the analytics or traffic dashboard you already use beside the relevant workspace, launch notes, tasks, and agent follow-ups.
Write the job once, attach its acceptance criteria, and launch Claude or Codex with the product context already connected.
Give every product a named, color-coded workspace and see which one is moving, blocked, or waiting for you before opening a pane.
Products spread across terminals, browsers, notes, and dashboards
Coding in one context and launching in another
An agent waiting while you work on a different product
Rebuilding context every time you switch products
Guessing which product needs attention next
One persistent workspace for every product
Build, launch, growth, and review work stay attached
Needs-action alerts show where a decision unlocks progress
Saved workspaces restore the complete working setup
Tasks, agent state, git signals, and browser context make it visible
CmdBrief is a local Rust app, not a cloud IDE, and it ships no usage analytics. Repositories and workspace state remain on your Mac. It contacts cmdbrief.com to check your subscription and sends redacted crash reports to Sentry (EU); agent CLIs and the optional voice assistant reach their providers only when you use them.
CmdBrief is for technical founders running Claude Code, Codex, or other agent CLIs across real products. Connect your first workspace and keep building, launching, growing, and reviewing a whole product portfolio from one place.
10 questions
CmdBrief is the native Mac workspace for technical founders running multiple products. It gives every product a persistent home for Claude Code, Codex, Grok CLI, Kimi CLI, OpenCode, and Cursor CLI, plus tasks, terminals, files, browser context, live agent status, usage meters, and session resume.
CmdBrief is built first for technical founders who personally build, launch, market, and review several products. It is most useful when switching between products means switching repositories, agent sessions, tasks, browser tools, and operating context.
Claude Code, Codex, Grok CLI, Kimi CLI, OpenCode, and Cursor CLI get the full supervision layer: live working, needs-action, completed, and failed states, notifications, and session resume. Usage meters cover Claude Code, Codex, Grok CLI, OpenCode Go plans, and Cursor CLI. Any other CLI can run in a regular terminal pane, but it does not report agent status.
Ask about agent status, workspaces, usage limits, approved terminal output, and project files after granting session-scoped workspace reading. Voice Agent can start or resume any supported agent and send dictated instructions; those instructions are translated into technical English before they reach an agent. Every state-changing action waits for local confirmation in CmdBrief.
OpenAI Realtime and xAI Grok are available. You can use a language supported by the selected provider and configure the assistant’s reply language. Language availability and accuracy vary by provider; each provider requires your API key and bills usage separately.
CmdBrief keeps product tasks, terminal sessions, source files, diffs, and browser panes in the same persistent workspace. You can coordinate launch checks, marketing deliverables, or traffic-driven implementation without pretending CmdBrief replaces your deployment, publishing, analytics, or search tools.
Not as a shipped native integration. Keep the analytics, search, CMS, or campaign tool you already use open in a browser pane, capture the next outcome as a task, and launch Claude Code, Codex, or another supported agent for bounded implementation work. Publication and measurement remain in the authoritative tool.
CmdBrief is local-first: repos, terminal history, and workspace settings remain on your Mac. Agents connect to their providers under your accounts, and account, billing, and optional voice features can use network services. Session browsing uses local metadata plus bounded local prompt and response previews.
Yes. CmdBrief is a native Apple-silicon Mac app requiring macOS 26+. Native pieces such as the menu-bar system monitor and notifications are part of why it is Mac-first.
Create your account with your email. We verify the email, then you can connect your workspace and start using CmdBrief on your products.