- Settings → About → "Start AgentHub automatically" (AutostartSettings.tsx,
GET/PUT /api/settings/autostart, server/lib/autostart.js). The switch is
the systemd --user unit's / LaunchAgent's enabled state: on writes a unit
for the running server when none exists and enables it, off disables it;
neither starts or stops the server. A foreign unit is never touched
(409 EFOREIGN)
- The plugin bootstrap no longer installs a unit on every install: it
installs or starts one only while the user has it enabled, and a rewrite
keeps the enabled state (writeUserService({ enable })). An install that
already had the unit enabled stays on. A stale-but-enabled unit beside a
live detached server is repointed, no longer started onto the bound port
- unitIsEnabled / setUnitEnabled in systemd-service.js and launchd-service.js
Also:
- fix(workspace): status bar no longer covers the chat box below md (#61)
18 KiB
Installation
A step-by-step guide to get AgentHub — the local control surface for Claude Code agents — up and running on your machine, with optional sections for importing history and running in a container.
Fastest path — install it as a Claude Code plugin
If all you want is a working dashboard, skip every step below. From Claude Code:
/plugin marketplace add https://git.smartgift.io.vn/Smartgift-AI/AgentHub.git
/plugin install agenthub@agenthub
The next session start installs the hooks, boots the server, puts the agenthub CLI
on PATH — no clone, no npm run setup, no
npm run install-hooks, no manual npm start. Run /agenthub-doctor to check the
result and /agenthub-open to build the UI and get the URL.
Requires Node >= 22.5 (a plugin install has no native better-sqlite3, so
the server stores data through node:sqlite). Full behavior, including the
uninstall cleanup, is in docs/PLUGINS.md.
Follow the steps below instead when you want a checkout — to develop the dashboard, run it from source, or stay on Node 20.
Requirements
| Requirement | Version | Notes |
|---|---|---|
| Node.js | 20+ (22+ recommended) | Required for server and client |
| npm | 9+ | Comes with Node.js |
| Claude Code | 2.x+ | Required for hook integration |
| Git | Any | Cloning the repository, and every lane operation (worktrees, branch/status reads) |
| tmux | 3.x (3.7+ for pane scrollbars) | Required for the run console — the browser terminal, agenthub workspace shell, and typing /reload-plugins into a live pane. See Step 0 |
| Codex CLI | Any | Optional — only for lanes whose tool is codex |
GitHub CLI (gh) |
Any | Optional — only for agenthub workspace ci status |
| Python | 3.6+ | Optional — only for scripts/expand-ts-module-docs.py (doc generation) |
Step 0 — System packages
Node and npm are not enough: git backs every lane operation, and tmux
backs the run console. The dashboard shells out to the tmux binary and has
no bundled fallback, so without it Start Run and agenthub workspace shell
fail while everything else keeps working.
sudo apt install git tmux # Debian / Ubuntu
sudo dnf install git tmux # Fedora / RHEL
brew install git tmux # macOS
sudo pacman -S git tmux # Arch
Windows has no native tmux. Run AgentHub under WSL2 and install it inside
the WSL distro — a tmux from Git Bash or MSYS is not what the server executes.
On a native Windows install, workspace terminals are unavailable, but Settings →
Agents still detects each AI CLI through PATHEXT (claude.exe, codex.cmd)
and runs a .cmd updater or npm through cmd.exe.
A terminal pane gets a draggable scrollbar only on tmux 3.7+ — older
versions work, just without it, and /agenthub-doctor warns. Homebrew ships the
latest (brew upgrade tmux); Ubuntu 24.04's apt has 3.4, so build the release:
sudo apt install build-essential pkg-config libevent-dev libncurses-dev bison
curl -fsSL https://github.com/tmux/tmux/releases/download/3.7c/tmux-3.7c.tar.gz | tar xz
cd tmux-3.7c && ./configure && make && sudo make install # into /usr/local/bin
A tmux server already running keeps its old version until every tmux session has closed (or the machine reboots) — only then do the scrollbars appear.
Check it:
tmux -V # e.g. tmux 3.4
git --version
node -v # v20.x / v22.x / v24.x
Once the server is up you can ask it directly — this is the same probe the dashboard uses:
curl -s localhost:4820/api/run/tmux # {"available":true}
No configuration is needed. AgentHub creates its own session per lane,
named agenthub-lane-<id>, and never reads your ~/.tmux.conf for anything it
depends on (mouse mode, the hidden status bar and the scrollbar are set on the
session explicitly). You can attach to the
exact same pane from a normal terminal:
agenthub workspace shell # from inside the lane's directory
tmux attach -t agenthub-lane-7 # or by name
tmux ls # every dashboard session is prefixed agenthub-lane-
Detaching is Ctrl-b d — it leaves the run alive, which is the whole point:
the browser terminal and your local terminal are two clients on one pane, and
restarting the dashboard never stops a run.
macOS
Status: code-audited, untested on a Mac. The Mac-specific paths below are covered by Linux-side unit tests only; the checklist at the end of this section is what a Mac user runs to confirm them.
brew install node git tmux # Node >= 22.5 for the plugin install (nvm works too)
echo 'export PATH="$HOME/.local/bin:$PATH"' >> ~/.zshrc # where the plugin puts `agenthub`
- Node: a plugin install needs >= 22.5 (it stores data through
node:sqlite); a checkout still runs on Node 20. Theagenthublauncher pins the node that installed it and falls back tonodeonPATHwhen abrew upgradeornvmswitch deletes that binary. - Starts at login once you turn it on (LaunchAgent). Off by default; with
Settings → About → Start AgentHub automatically on, the plugin bootstrap
installs
~/Library/LaunchAgents/io.smartgift.agenthub.plist(a checkout: the same switch, oragenthub service install), so the dashboard comes back after every login with no Claude Code session, and launchd restarts it if it dies. Lane panes survive a restart of the server. Its output goes to~/Library/Logs/agenthub.log. A LaunchAgent needs a login at the Mac itself; over SSH only, or whenlaunchctlfails, the server is a detached process started by the next Claude Code session (within about 10 s) or byagenthub start, as before. A discovery entry left over from before the reboot is ignored, so a recycled pid can never pass for a live dashboard. - Started from a Claude Code session. The server drops the variables that
describe that session (
CLAUDECODE,CLAUDE_CODE_SESSION_ID,CLAUDE_PLUGIN_ROOT, …) before anything else runs, so lane panes do not inherit them. Your own configuration (CLAUDE_CONFIG_DIR,CLAUDE_CODE_USE_BEDROCK,ANTHROPIC_*) passes through. A tmux server that an older build started keeps its dirty global env until it exits (tmux kill-serveronce, with no run you care about open). - Paths. A lane's directory is stored by its real path, so a typed
/tmp/xbecomes/private/tmp/xand a differently-cased path on a case-insensitive volume becomes the one on disk — that is what sessions report, so they bind. Lanes created before this change keep the string they were created with. - Account switching uses the login Keychain item
Claude Code-credentials; macOS may ask once for permission the first time. - Option key in the browser terminal types special characters (it is not sent as Meta).
Mac checklist — please run it and report the result on issue #22:
- Fresh plugin install from a Terminal
claudesession: the dashboard answers onhttp://localhost:4820,~/Library/LaunchAgents/io.smartgift.agenthub.plistexists, and/agenthub-doctor's launchd row says the LaunchAgent is current and running. tmux show-environment -g | grep -i claudeprints nothing.- Add a Repo lane by typing its path in a different case: sessions bind to it, and
agenthub stage planinside it moves the map. - Start Run in the browser: the terminal attaches, typing works, an image paste into
claudeworks, andexitreturns to the workspace home. - Changes panel: diff, stage and commit on a lane.
- Switch profile and check
claude /statusshows the new account. Repeat with a non-defaultCLAUDE_HOME, which also checks the Keychain service name. brew upgrade node && brew cleanup, thenagenthub statusstill works.- Reboot and log in WITHOUT opening
claude: the dashboard answers within about 10 s (agenthub service statusshowsstate = running). kill <server pid>(fromagenthub status): it is back within a few seconds.- With a lane run open,
/agenthub-updatecompletes, the server restarts on the new version, the plist names the new cache directory, and the run's terminal is still alive. agenthub service uninstall: the plist is gone and nothing restarts the server after akill.
Step 1 — Clone the repository
git clone https://git.smartgift.io.vn/Smartgift-AI/AgentHub.git
cd AgentHub
Step 2 — Install dependencies
npm run setup
This installs all server and client dependencies, plus the VS Code extension, and links the agenthub CLI.
A plain root install already covers server and client — a postinstall hook installs the client dependencies automatically, so this alone is enough to build and run the dashboard:
npm install
npm run setup additionally installs the VS Code extension and links the agenthub CLI. (If you install with --ignore-scripts, the postinstall hook is skipped — run cd client && npm install manually in that case.)
Or via Makefile:
make setup
Step 3 — Start the dashboard
npm run dev
This starts two processes concurrently:
| Process | URL | Description |
|---|---|---|
| Express server | http://localhost:4820 | API, WebSocket, SQLite |
| Vite dev server | http://localhost:5173 | React frontend with HMR |
Open http://localhost:5173 in your browser.
Tip
When you run the dashboard directly on the host with
npm run devornpm start, the server automatically writes the Claude Code hook configuration to~/.claude/settings.json— unless theagenthubplugin is enabled there, whose own hooks already run; then it removes any copies a checkout wrote instead, since two sets post every event twice. If you run the dashboard inside a container, install hooks from the host withnpm run install-hooksinstead — a hook path written from inside a container points at the container filesystem.
Step 4 — Start a Claude Code session
Start a new Claude Code session from any directory after the dashboard server is running. The hooks will fire automatically and your sessions, agents, and events will appear in real-time.
# In a separate terminal, from any project directory:
claude
Verification
After starting a Claude Code session, you should see:
- Sessions page — your session listed with status
Waiting(a fresh CLI sitting at the prompt) orActive(mid-turn) - Agent Board — a
Main Agentcard in theWaitingcolumn until you type your first message; it flips toWorkingonUserPromptSubmit/PreToolUseand back toWaitingafter eachStop - Activity Feed — events streaming in as Claude Code uses tools
- Dashboard — stats updating in real-time
- Settings page — data export and cleanup tools on the Dashboard tab; Claude Code's model pricing rules, hook configuration status and CLAUDE_HOME on Settings → Claude Code → Overview
If nothing appears after 30 seconds, see SETUP.md.
PWA install (optional)
The dashboard is a Progressive Web App. After opening it in a supported browser (Chrome, Edge, Firefox), you can install it to your dock / home screen:
- Look for the install icon (⊕) in the browser address bar, or open the browser menu → "Install app"
- Once installed, the dashboard launches in its own window with no browser chrome
- Offline support: previously visited pages and assets are served from the Service Worker cache when the network is unavailable
Step 5 — (Optional) Import existing Claude Code history
The server automatically imports every session under
~/.claude/projects/ on startup, so if you've used Claude Code on this
machine before, your historical sessions, agents, events, token counts,
and cost totals should already be visible in the Sessions list.
To bring in history from another machine, a backup, or a .tar.gz
archive a teammate sent you, use Settings → Import History in the
UI. It supports three modes:
flowchart LR
UI["Settings →<br/>Import History"] --> M1["Rescan default folder<br/>~/.claude/projects"]
UI --> M2["Scan a folder<br/>any absolute path"]
UI --> M3["Upload files<br/>.jsonl / .zip / .tar.gz / .gz"]
M1 --> P["Same parser as live<br/>hook ingestion"]
M2 --> P
M3 --> P
P --> DB[("SQLite<br/>sessions + tokens + cost")]
style UI fill:#a855f7,stroke:#c084fc,color:#fff
style P fill:#f59e0b,stroke:#fbbf24,color:#000
style DB fill:#10b981,stroke:#34d399,color:#fff
Re-imports are idempotent: sessions are deduplicated by UUID and compaction baselines preserve pre-compaction token totals, so running the importer twice never double-counts tokens or cost.
Verify it worked by opening the Analytics page and checking that per-model token totals and estimated cost look correct. Full walkthrough with per-OS archive commands in SETUP.md → Importing existing Claude Code history.
Optional: tune import limits
If you regularly import very large archives, these environment variables can be raised (the defaults are generous for typical use):
| Variable | Default | Purpose |
|---|---|---|
AGENTHUB_IMPORT_MAX_BYTES |
1 GB | Maximum size per uploaded file |
AGENTHUB_IMPORT_MAX_FILES |
2000 | Maximum files per upload request |
AGENTHUB_IMPORT_MAX_EXTRACT_BYTES |
4 GB | Ceiling on uncompressed bytes per archive (zip-bomb guard) |
Set them before npm run dev or npm start:
AGENTHUB_IMPORT_MAX_EXTRACT_BYTES=17179869184 npm start # allow 16 GB extraction
Production mode
To run as a single process serving the built client:
npm run build # Build the React client
npm start # Start Express serving client/dist on port 4820
Open http://localhost:4820 in your browser.
npm start (and agenthub start, its backgrounded form) only run while that
process is alive — a reboot loses it. Run the server as a systemd --user
service (Linux) or a LaunchAgent (macOS) instead so it comes back on its own
after every reboot, with no manual step and no Claude Code (or any other tool)
needed to trigger it:
agenthub service install # Linux: ~/.config/systemd/user/agenthub.service
# macOS: ~/Library/LaunchAgents/io.smartgift.agenthub.plist
agenthub service status # systemctl --user status / launchctl print
agenthub service uninstall # disable and remove it
See README.md → Run as a service
and, for plugin installs, docs/PLUGINS.md → Auto-start on boot.
Optional: Agent extension packs
This repository includes extension packs for both Claude Code and Codex.
- Claude Code loads project extensions from:
CLAUDE.md.claude/rules/skills/(shipped as theagenthubplugin's skills)
- Codex project packs live under
.codex/:AGENTS.md.codex/rules/.codex/agents/.codex/skills/
See .codex/README.md for Codex extension details.
Troubleshooting
npm run setup shows better-sqlite3 errors
This is expected and non-fatal. better-sqlite3 is a native C++ module listed as an optional dependency. If prebuilt binaries are not available for your Node version or platform, npm will print gyp/compilation errors but still complete successfully.
At runtime the server uses this fallback chain:
better-sqlite3— used when prebuilt binaries are available (Node 20/22/24 on Windows x64, macOS arm64/x64, Linux x64/arm64)node:sqlite— Node.js built-in SQLite module, used automatically on Node 22+ whenbetter-sqlite3is unavailable
If you see an error box at startup saying "SQLite backend not available", either:
-
Upgrade to Node.js 22+ (recommended — zero native dependencies needed), or
-
Install build tools so
better-sqlite3can compile from source:- Windows:
npm install -g windows-build-toolsor install Visual Studio Build Tools with the C++ workload - macOS:
xcode-select --install - Linux:
sudo apt install python3 make g++(Debian/Ubuntu) or equivalent
Then run:
npm rebuild better-sqlite3 - Windows:
npm run dev fails immediately
Ensure both server and client dependencies are installed:
npm run setup
If the error mentions a missing module like express or react, dependencies may be incomplete. Delete node_modules in both root and client/, then re-run setup:
rm -rf node_modules client/node_modules
npm run setup
Server starts but client shows a blank page
The Vite dev server and Express server run on different ports. Make sure both are running (npm run dev starts both). Open http://localhost:5173, not http://localhost:4820, during development.
"Start Run" fails, or the terminal never attaches
The run console is tmux-backed. Check, in this order:
tmux -V # not installed → see Step 0
curl -s localhost:4820/api/run/tmux # {"available":false} → the SERVER cannot see tmux
curl -s "localhost:4820/api/run/binary?tool=claude" # {"found":true,"path":"…"}
{"available":false} while your own shell finds tmux means the server runs with
a different PATH — most often a systemd --user service, which does not read
your shell profile. Fix it by installing tmux system-wide (/usr/bin/tmux)
rather than into a shell-managed prefix, then agenthub service uninstall && agenthub service install.
The same PATH rule applies to claude and codex: a run starts the binary by
name, so a claude installed only through an interactive shell's profile is
invisible to a service-managed server. {"found":false} for the lane's tool is
that case.
No sessions appearing after starting Claude Code
See SETUP.md — Troubleshooting for detailed hook debugging steps.
Ports
| Service | Default | Override |
|---|---|---|
| Dashboard server | 4820 |
DASHBOARD_PORT=xxxx npm run dev |
| Client dev server | 5173 |
Edit client/vite.config.ts |