mulmoterminal

Run multiple Claude Code and Codex sessions in parallel — and see which one needs you.

A browser terminal for parallel AI coding agents: several Claude Code and Codex sessions side by side, each in its own cell, with the one that needs you marked in colour. Vibe coding with a single agent needs nothing but a shell — this is for when you run several and lose track of which is waiting. Sessions survive a reload (tmux), work isolates in git worktrees, and a phone push reaches you when a turn finishes.

📖 Documentation — receptron.github.io/mulmoterminal

• User guide: English — the grid view, everyday workflows, the full feature list, configuration, and mobile push notifications.
• ユーザーガイド: 日本語 — グリッドの使い方・日々のワークフロー・機能一覧・設定・スマホ通知の設定はこちら。
• Updates / アップデート情報: new releases and features are announced in Japanese on X — 新バージョンや新機能のお知らせは X の Singularity Society (@SingularitySoci) で。

MulmoTerminal turns Claude Code (and OpenAI's Codex) into a parallel, observable workspace: many agent sessions at once in a grid, each one color-coded so you see at a glance which are working, which need you, and which are done — plus rich GUI output, git worktrees with one-click PRs, cost readouts, and a ping to your phone when a task finishes. One npx command, no Electron, no config.

npx mulmoterminal@latest        # starts on http://localhost:34567 and opens your browser

Something looks wrong? Type /mulmoterminal-bug-report in any MulmoTerminal session. The bundled skill hears the symptom out, checks your real config, schema and version to see whether the behaviour is configuration or by design, searches the existing issues — and only helps you file one if none of that explains it, with the environment collected and secrets masked. Getting you unstuck is the goal; an issue is what is left when the first three steps fail.

The grid is a cockpit for parallel agents — here, four live Claude sessions, each in its own color-coded project. Every cell's header carries what you need to triage at a glance: model · context %, token counts (⇡in ⇣out), the git branch / changes chip, and an AI summary of what the agent is doing. A cell's border color signals state — working / done (blue), needs-you (amber — e.g. waiting on a permission), idle — with an attention chime so a stuck cell off-screen still pulls you back. Supervise many; only step in where you're called.

Why you'll want it

• See every agent at once. A grid of live sessions, each cell color-coded by state — working (blue), blocked / needs a permission (amber), done, unreviewed (blue), idle — with an attention chime and a toolbar tally, so an off-screen agent that's stuck never slips past you. Stop babysitting one terminal; supervise ten. Zoom into one and the cockpit roster keeps everyone else in view — one text row per session with its AI summary, last prompt, latest reply, and the branch's PR phase (draft / CI fail / ready / merged).
• A GUI for your agents, not just a terminal. Beside the terminal, a Canvas panel renders what an agent produces over MCP — documents, forms, charts, generated images, HTML, collection cards — each drawn by its own plugin. The agent doesn't just print text; it hands you an interface.
• Get pulled back from anywhere. A finished — or input-waiting — task sends a Web Push to your phone, and the RemoteHost companion lets you watch sessions and answer with a tap (yes / no / continue) from the phone itself — walk away, get pinged, jump back in.
• Nothing is lost on a restart. With tmux, every session survives a server crash, restart, or node --watch reload — a mid-turn agent, a long build, a dev server all keep running and reattach when you come back.
• Ship without leaving the grid. Each repo cell shows a git branch chip, isolates work in a one-click git worktree, opens a diff panel, and does commit / push / open PR — so several agents can work the same repo without colliding.
• Know what it's costing. Per-session context %, token, and estimated $ readouts, an activity timeline of tool calls, and AI-summarized cell titles and command-output explanations — so a wall of parallel agents stays legible.
• Make it yours. Per-directory themes, colors, and name badges (prod in red, staging in amber), a configurable header (buttons + info chips), custom attention sounds, and Run / Skill menus to launch a project's scripts and .claude/skills right inside a cell.

Zoomed in, the cockpit roster replaces thumbnails with information: every session as a text row — directory, AI summary, your last prompt, the agent's latest reply, a status word, and the branch's PR phase badge. A row whose agent is waiting on you rings amber and blinks; one that has merely finished rings green and stays still (Settings → Waiting rows turns the movement off). Click a row to swap the enlarged terminal.

What it is, under the hood

Each session runs as a real PTY on the server (the agent CLI in a pseudo-terminal) and is streamed to an xterm.js terminal in the browser over a WebSocket. A sidebar lists every session for the project and reflects, in real time, which are working (the agent is thinking, a spinner), which are waiting on you (a permission prompt or a question — an amber dot; nothing proceeds until you answer) and which are finished with output you haven't seen (a green dot) — driven by Claude/Codex activity hooks the server injects per spawn. The horizontal tab bar carries the same two dots.

Besides the grid there's a single view for focusing on one agent: the conversation/terminal on the left, and a GUI panel ("Canvas") on the right where the agent's tool calls render as documents, forms, charts, images, and HTML — not just printed text. Switch between the two with the chat / grid icons in the toolbar. The app opens on the grid (/); the single view has its own URL, /chat, so you can bookmark either.

Inserting a file path — like a native terminal, you can put a file's absolute path into the prompt: drag a file onto the terminal, or click the file button in the terminal header, which asks the local server to open the OS file dialog and inserts the chosen path. The path is inserted at the cursor — it is not submitted, so you can review it first.

A drag inserts the file's own path where the browser exposes one via file:// (Firefox/Safari), so editing it afterwards edits the file you dropped. Where the browser withholds it — Chrome, and every browser when MulmoTerminal is open from another machine, where a local path would name nothing on the host — the file's bytes are sent instead, saved to a private per-session directory under the OS temp dir, and that path is inserted. The session is granted that directory at launch (Claude Code's --add-dir, bind-mounted in the sandbox too), so the agent reads it without a permission prompt; the copies are removed when the session ends, and any left by a crash are swept at the next start. Up to 110 MiB per file — the same ceiling as a phone attachment. A session already running when you upgrade was launched without that grant, so drops into it still prompt; new sessions don't.

Pasting a screenshot — take a screenshot and paste it straight into the terminal (Cmd/Ctrl+V). The image is saved to the session's own drop directory — the same place a dropped file goes, with the same grant, the same 110 MiB ceiling and the same cleanup when the session ends — and its absolute path is inserted at the cursor, so the agent can read it. Unlike a drop, this does not need the browser to expose a path — the bytes are on the clipboard — so it also covers Chrome, where dropping a file cannot insert a path. It works wherever the browser puts the image on the clipboard as image/png, image/jpeg, image/gif, or image/webp. Anything else is left to the terminal's own paste handling, exactly as before — including a paste that carries plain text next to the image, which copying from a web page usually does, so that pasting text keeps working.

Clicking a file path — the other direction. A path an agent prints becomes a link, and what it opens is chosen by its extension, so each kind arrives as the thing it is rather than as bytes (files within the session's working directory only):

A clicked …	opens as
.md .markdown	rendered markdown in a new tab — the same sandboxed …/md HTML the Files preview uses. It follows your system light/dark setting, since under the sandbox CSP it can't ask the app which theme is on
.json	indented in a new tab (Chrome and Safari otherwise show one long line)
.csv .tsv	a table in a new tab, with a sticky header that scrolls inside its own box
source, config, logs, and .txt — 46 extensions	the app's own Files view (/files?path=), where CodeMirror highlights it, the tree is right there, and it can be edited
everything else — images, PDF, SVG, HTML, video	raw bytes in a new tab, which the browser renders better than an editor would

While a grid cell is enlarged, the Files pane takes the click first — every row above except the last one, since the pane is the same editor plus a Markdown preview. The file opens beside the terminal that printed it, and the pane opens itself if it was closed. It declines, leaving the routing above untouched, when nothing is enlarged, when the path is not under that cell's own directory (the pane cannot walk above its root), or for the raw-bytes row, where it would only show an empty editor.

Highlighting in the Files view covers the JS/TS family, JSON and Markdown (the modes cmEditor.ts bundles); other languages open as plain text.

This set is deliberately asymmetric with the set the server serves as viewable text — .md goes to the rendered viewer rather than the Files view, .txt does the opposite, and dotfiles are server-only. The 45 extensions both sides agree on live in common/sourceExtensions.ts, each side adds its own extras, and test/common/sourceExtensions.spec.ts pins the asymmetry so it isn't "fixed" into symmetry.

Changing this? The routing table is ROUTE_BY_EXTENSION / IN_APP_EXTENSIONS in src/composables/terminalFilePathLinkProvider.ts. Update this section, the docs/guide/{en,ja}/features.md row, and the link table in docs/terminal-notes.md together — all three went stale once already (#834).

---

What people say after switching

These are experiences reported by users who moved over from an IDE or a split terminal — not benchmarks, and not claims we measured. Your setup may differ.

"It stopped eating my memory"

Keeping several agents apart by opening several IDE windows is expensive: each one brings its own editor, language server, extensions and file watchers. One user reported a 64 GB machine stuttering under that load, and running smoothly after moving over — here the agents are PTYs on a server and the UI is browser tabs.

"I stopped answering the wrong agent"

Six panes of scrolling text look identical. Users have described typing a reply into another agent's terminal, and losing track of what they had asked in the first place. As one put it, the windows all look the same, so switching between them costs time just to work out what you are looking at.

The problem isn't attention — it's that N identical panes means holding N contexts in your head. Colour-coded state, a name badge and a per-directory colour move that onto the screen instead.

"Watching many and reading one stopped being a trade-off"

Splitting a terminal six ways leaves every pane too small to read a long answer without constant scrolling and resizing — one user described exactly that with a 4,000-character reply. So you quietly accept worse reading every time you add an agent.

Grid ↔ enlarge removes that. Watch all of them, then blow one up and read it properly — the cockpit roster keeps the rest in view as text while you do.

"My existing sessions came with me"

Sessions resume as-is — same claude --resume, same transcripts. Point it at a directory you already work in and your history is there. Nothing to migrate, nothing to redo. One user said this alone made the switch worth it, having previously lost context to killed sessions.

---

You don't need ten agents for this to pay off. Users have reported the switch being worth it at one to three parallel sessions. The wins above are about not losing track, not about running more.

---

Install & run

Needs Node ≥ 22.9, plus these CLIs on your PATH:

	Tool	What it gives you	Install
Required	claude	every Claude session — this app is a cockpit for it	npm i -g @anthropic-ai/claude-code, then run claude once to log in
Required	git	worktree isolation, each cell's branch / unsaved-dot / diff readout, the PR footer	brew install git · sudo apt install git · sudo dnf install git · Windows: git-scm.com
Required	gh	the cross-repo PRs & Issues view and one-click PR creation — it uses your gh login, so no token is stored	cli.github.com, then gh auth login
Recommended	tmux	session persistence — terminals survive a server restart	brew install tmux · sudo apt install tmux · sudo dnf install tmux · no native Windows build (falls back to plain PTYs)
Optional	codex	Codex sessions in a cell, alongside Claude	npm i -g @openai/codex
Optional	docker	the experimental Docker sandbox	docs.docker.com
Optional	ffmpeg	video rendering from the mulmo-script panel (its plugin ships enabled)	brew install ffmpeg · sudo apt install ffmpeg · sudo dnf install ffmpeg
Optional	ollama	claude-ollama — Claude Code against a fully local model	ollama.com/download

The server starts without any of the non-required rows; you just lose that row's feature, and the header/panel for it says so. git and gh are marked required because losing them costs whole views rather than one button. npx mulmoterminal@latest init (below) reports which of these it can find.

npx mulmoterminal@latest           # start on http://localhost:34567 and open the browser
# or install globally:
npm install -g mulmoterminal
mulmoterminal

First-run setup (optional). npx mulmoterminal@latest init checks your environment (Node ≥ 22.9 and every CLI in the table above), seeds the launcher's directory presets from the projects in your Claude Code history, and writes ~/.mulmoterminal/config.json. It's idempotent — re-run it any time to refresh the presets; it overwrites the managed parts and keeps your other settings. When claude is installed it can hand off to the /mulmoterminal-config skill for interactive tweaks — it routes to the one that owns what you want to change. Once the app is up you can also reach them from Settings: each section that a skill can write ends in a button that starts that skill in a new session, which is how the settings with no UI (a theme of your own, keymap) get written without hand-editing JSON.

Google account (optional). Link a Google account to enable the chat's google tool and the phone's google.calendar.* commands: read/create events on any calendar (not just your primary), list the calendars you've subscribed to, and read the colour palettes. Sign in from Settings → Google account, or run npx mulmoterminal@latest google login — the CLI is the fallback for when you're driving MulmoTerminal from another machine, since consent finishes on a loopback listener and needs a browser on the host. Either way it needs a Desktop OAuth client JSON saved as ~/.secrets/client_secret_*.json; the refresh token lands in ~/.config/mulmo/google-token.json and is shared with MulmoClaude, so one link per machine covers both apps.

Local models (optional). The package also ships claude-ollama — a one-command launcher that runs Claude Code fully locally against an Ollama model (no cloud, no API key). It starts a large-context Ollama server and launches claude with a minimal system prompt so small models aren't drowned:

ollama pull qwen3:4b
npx -p mulmoterminal claude-ollama qwen3:4b   # or, if installed globally: claude-ollama qwen3:4b

See Local models with claude-ollama for the details and model notes.

Already linked before the calendar-list / colour features? They need a read scope your existing link doesn't have, so listCalendars (and, in practice, colors) fail with an insufficient-scope 403 until you re-authorize: Settings → Google account → Unlink, then sign in again (or re-run google login). Reading/creating events on your primary calendar keeps working without re-linking.

A global install isn't auto-updated, so on startup MulmoTerminal checks npm and prints a one-line notice when a newer version is available — and the web toolbar shows a clickable update badge with the exact command for your install (npm i -g mulmoterminal, or git pull for a clone). Disable with MULMOTERMINAL_NO_UPDATE_CHECK=1 (or NO_UPDATE_NOTIFIER=1).

Options: --cwd <dir> (working directory — relative paths allowed; defaults to the directory you run the command from), --port <n> (default 34567), --no-open, --version, --help.

npx mulmoterminal@latest --cwd ./my-project   # work in a specific directory

The published package ships the server (run via tsx) plus the pre-built web UI; npx mulmoterminal@latest checks for the claude CLI, picks a free port, starts the server, and opens the browser. For local development from a clone, see Running.

Won't start with ERR_MODULE_NOT_FOUND? If a first npx run was interrupted, a half-unpacked ~/.npm/_npx/<hash> cache can remain and a later run fails at startup — a corrupted npx cache, not a bug in the published package. The launcher detects it and prints the exact, OS-appropriate removal command; run that, then npx mulmoterminal@latest again.

---

Contents

• Architecture
• Why a PTY?
• Agents: Claude & Codex
• Session persistence (tmux)
• Docker sandbox (experimental, single view)
• Tech stack
• Configuration
• Running
• Scripts (Run menu)
• Skills (Skill menu)
• Files view (browse & edit)
• Git worktrees & pull requests
• Cost & token usage
• Wiki, Collections & the GUI panel
• More features
• Server API specification
  • HTTP: GET /api/sessions
  • HTTP: GET /api/scripts
  • HTTP: POST /api/command/summarize
  • HTTP: POST /api/hook
  • More HTTP endpoints
  • WebSocket: /ws (terminal)
  • More WebSocket endpoints
  • WebSocket: /ws/run (command terminal)
  • Socket.IO: /ws/pubsub (activity pub/sub)
• Session model
• Session lifecycle
• Claude hook injection
• Closing summary
• Session discovery & titles
• Project structure
• Testing

---

Architecture

┌──────────────────────────────────────┐         ┌─────────────────────────────────────────────┐
│ Browser (Vue 3 + xterm.js)            │         │ Server (Express + Node)                       │
│                                       │         │                                               │
│  Sidebar.vue ──subscribe("sessions")──┼──SIO───►│  socket.io  /ws/pubsub   ── publish ──┐       │
│      ▲  refetch on any push           │         │                                       │       │
│      └──── GET /api/sessions ─────────┼──HTTP──►│  Express   /api/sessions              │       │
│                                       │         │            /api/hook  ◄──curl── hooks │       │
│  Terminal.vue ── ws JSON msgs ────────┼──WS────►│  ws        /ws  ──► node-pty ─► `claude`──hooks┘
│      (input / resize / output)        │         │                     (one PTY per session)     │
└──────────────────────────────────────┘         └─────────────────────────────────────────────┘

• Terminal I/O flows over a raw WebSocket (/ws), one PTY per session.
• Session list is fetched over HTTP (/api/sessions).
• Live activity is pushed over a Socket.IO pub/sub channel (/ws/pubsub); the server learns of activity from Claude hooks that POST to /api/hook.
• Other terminals run on their own raw WebSockets: Codex sessions on /ws/codex, persistent launch commands on /ws/launch, and one-off script commands (yarn dev, tests, …) on /ws/run. Only Claude/Codex are agent sessions with hooks; see Agents: Claude & Codex and Scripts (Run menu).
• In dev (yarn dev) the Vite dev server runs on its own port (CLIENT_PORT, default 6856) and proxies /ws (a prefix covering /ws/codex, /ws/launch, and /ws/run), /ws/pubsub, /api, and /artifacts to the backend (PORT, default 34567) — so you open the Vite port (e.g. http://localhost:6856). In production the backend serves the built client from dist/ on PORT, and you open that.

---

Why a PTY?

Claude Code's interactive mode renders its UI with Ink (a React-based TUI framework), which requires a real TTY to be attached. A plain child_process.spawn() provides no TTY, so interactive Claude won't start (it stays silent). node-pty allocates a real pseudo-terminal at the OS level, so from Claude's point of view it's running in an ordinary terminal — full TUI rendering, cursor movement, colors, and tool-approval prompts all work. We don't use -p/headless mode or the Agent SDK; we drive the real interactive CLI and relay its TTY over the WebSocket.

macOS note: node-pty's bundled spawn-helper binary ships without the execute bit (mode 644), which causes a posix_spawnp failed error. The postinstall script (server/fix-pty-perms.js) fixes it to 755 automatically.

---

Agents: Claude, Codex & Antigravity

MulmoTerminal drives interactive coding-agent CLIs, not just Claude. An AgentAdapter seam abstracts the per-agent bits (which binary to spawn, how it resumes) so the PTY, grid, persistence, and GUI-panel plumbing stay shared. Three adapters ship today — Claude Code (the default), Codex, and Antigravity (agy).

• Claude — spawned as claude (override with CLAUDE_BIN). The server passes --session-id <uuid>, so it knows the live session's id even before its transcript file exists, and injects activity hooks + the GUI MCP per spawn (see Claude hook injection) plus the closing summary instruction.

• Codex — spawned as codex (override with CODEX_BIN; CODEX_MODEL sets --model). Codex runs on its own WebSocket (/ws/codex) and its sessions appear in the sidebar next to Claude's. Because Codex only mints its rollout id after the first turn, the server watches ~/.codex/sessions/**/rollout-*.jsonl (home overridable via CODEX_HOME) and maps the new rollout to the session — attributed only when it's unambiguous, never by "newest wins". Resume reattaches a live PTY, adopts a surviving tmux session, or cold-resumes the rollout id.

• Antigravity — spawned as agy (override with ANTIGRAVITY_BIN; ANTIGRAVITY_MODEL sets --model). Antigravity runs on its own WebSocket (/ws/antigravity). Like Codex it mints its own conversation id, so the server watches ~/.gemini/antigravity-cli/brain/ (home overridable via ANTIGRAVITY_HOME) for the directory the new conversation creates — attributed only when unambiguous — and cold-resumes it with --conversation <id>. That mapping is appended to ~/.mulmoterminal/antigravity-conversations.jsonl, so a conversation is still resumable after the server restarts.

  Its GUI tools work differently, because agy takes no MCP flag: it reads its servers from .agents/mcp_config.json in the working directory. MulmoTerminal writes that file from the directory's Canvas switches — the same switches Claude's cells read — so one switch serves every agent, and rewrites it whenever a switch flips or an agy session starts. Servers in it that MulmoTerminal did not write are left alone, the file is removed once no group is on, and it is kept out of your git status through .git/info/exclude — a local switch on a local machine, so it never reaches a diff or your team. The entry runs server/mcp/bridge.mjs, a stdio-to-HTTP shim onto the same in-process GUI MCP server the other agents call. The session id is never written into that file — it is per directory and shared by every session running there — and reaches the bridge through the agy process's own environment instead.

  The Docker sandbox does NOT cover agy: it stays claude-only until buildDockerRunArgs is generalized (see plans/feat-multi-agent-support.md, PR#5). agy also ships as a standalone binary rather than an npm package, so the sandbox image has nothing to install.

Choosing an agent. The single view has a New Codex session button; each grid cell's launch form carries a Claude / Codex / Antigravity / Shell toggle, and the Collections browser a Claude / Codex / Antigravity one (your choice is remembered). Shell is not an agent: it runs your OS default shell ($SHELL, or /bin/sh) in the chosen directory, with nothing to install and nothing to configure. It starts a launcher cell, so it has no model, no MCP registration, and no worktree — those rows disappear while it is picked.

Other models. Claude Code can run against any Anthropic-compatible backend (OpenRouter, Moonshot, a LiteLLM gateway). Backends are listed in ~/.mulmoterminal/config.json under providers, and their keys are read from the server's environment — never from a file the app serves. A directory sets its default in .mulmoterminal.json (provider / model), and each grid cell's launch form has a MODEL select that overrides it for one session, listing ~27 curated models with the measured pass rate of a real tool-using task beside each. A provider whose token can't be resolved refuses to start rather than falling back to Anthropic, and providers can't be combined with the Docker sandbox. Full walkthrough — setup, the measured model list, adding your own models, troubleshooting: Using another model via OpenRouter.

Skills for Codex. Codex has no /<slug> slash commands, so on session setup MulmoTerminal mirrors the workspace's .claude/skills into ~/.codex/skills (each mirrored directory carries a .mt-mirror marker so a re-sync overwrites what MulmoTerminal owns and never clobbers Codex's own skills), and rewrites a collection's /<slug> … seed into a plain Use the "<slug>" skill. instruction. The same skills Claude uses then show up for Codex, loaded by description.

---

Session persistence (tmux)

If tmux is installed, MulmoTerminal runs each Claude session and launcher inside a tmux session, so a server crash or restart doesn't kill your terminals — the processes keep running and reattach when the server comes back (like screen/tmux). A long build, a dev server, or a mid-turn Claude session all survive node --watch reloads and crashes. It uses its own tmux server (-L mulmoterminal) and config, so it never touches your personal tmux sessions or keybindings.

No tmux? No problem — terminals fall back to plain (non-persistent) PTYs, exactly as before. An explicit close (a cell's ✕) ends the tmux session; a machine reboot does not survive (tmux itself is gone). Command-cell scripts are ephemeral and not persisted.

Installing tmux (optional):

brew install tmux            # macOS (Homebrew)
sudo apt install tmux        # Debian / Ubuntu
sudo dnf install tmux        # Fedora

On Windows there's no native tmux, so sessions use the non-persistent fallback — run the server under WSL if you want persistence. Nothing else is required: MulmoTerminal detects tmux on PATH at startup and uses it automatically when present.

---

Docker sandbox (experimental, single view)

Set MULMOTERMINAL_SANDBOX=1 (and have Docker running) to run the single-view Claude session inside a container instead of on the host, while Claude still reaches the app's GUI MCP + activity hooks over host.docker.internal. The mulmoterminal-sandbox image is built automatically on first launch from the shipped Dockerfile.sandbox (~1 min, once; rebuilt only when that file changes). Override the name with MULMOTERMINAL_SANDBOX_IMAGE. If the image can't be built (e.g. Docker down), the session falls back to the host spawn — no cryptic failure.

This contains Claude — it can't reach the host filesystem outside the mounts, host processes, or arbitrary host ports. It is not full isolation: the workspace and ~/.claude are bind-mounted read-write by design (so Claude edits your project, and transcripts interoperate with host sessions), so those specific paths stay mutable from inside. The sandbox is non-persistent (the container is --rm, tied to the session), opt-in and single-view only — the grid keeps its host + tmux path, and with the flag unset (or Docker unavailable) everything runs on the host exactly as before. macOS only for now — on Linux (bind-mount uid ownership) and Windows (host paths aren't valid Linux container paths) it falls back to the host spawn; both are follow-ups. Adding arbitrary user MCP servers to the sandbox is in progress (see #202).

Authentication (macOS). Claude's live login token lives in the macOS Keychain, which the container can't read (mounting ~/.claude alone isn't enough — its .credentials.json is often absent or stale). On each sandbox spawn MulmoTerminal exports the current credential to a per-session ~/.mulmoterminal/sandbox/creds-<id>.json (mode 0600, removed when the session ends) and mounts it read-only over the container's ~/.claude/.credentials.json; your host ~/.claude is never modified. If you've never logged in on the host, run claude once first — otherwise the server logs a warning and the container shows "Not logged in".

Host credentials (opt-in). By default the sandbox has no host credentials. To let the sandboxed Claude use gh/git, set SANDBOX_MOUNT_CONFIGS=gh,gitconfig — a fixed allowlist (you pick names, never arbitrary paths): gh mounts ~/.config/gh read-only and passes a GH_TOKEN (from gh auth token, since macOS keeps it in the Keychain), and gitconfig mounts ~/.gitconfig read-only. Set SANDBOX_SSH_AGENT_FORWARD=1 to forward the SSH agent socket (the keys never enter the container). Both are read only when building the sandbox spawn, so they have no effect unless MULMOTERMINAL_SANDBOX is on.

---

Tech stack

Layer	Technology
Frontend	Vue 3 (<script setup> + TypeScript), Vue Router, Vite, xterm.js (@xterm/*), CodeMirror 6, socket.io-client
Backend	Node (ESM, TypeScript run via tsx), Express 5, ws (terminal WebSocket), node-pty, socket.io, @modelcontextprotocol/sdk (in-process GUI MCP)
Plugins	GUI-protocol Vue plugins (@mulmoclaude/*, @mulmochat-plugin/*): markdown, form, image, chart, HTML, collection, accounting, mulmoscript (MulmoCast video/slides), google
Tests	Vitest + @vue/test-utils + jsdom

Requires Node ≥ 22.9 (uses node --env-file-if-exists) and the claude CLI on PATH.

---

Configuration

The server is configured entirely through environment variables, optionally loaded from a .env file. npx mulmoterminal@latest reads the .env in the directory you run it from; the npm scripts read the one in the repo root. The .env is optional — every variable below has a default, so the server runs without one.

A variable already set in your shell wins over the same name in .env, so adding a file never overrides what you exported. The server's environment is inherited by every terminal it starts, so anything in .env is also visible to the claude / codex sessions themselves.

Variable	Default	Description
PORT	34567	Backend HTTP/WebSocket port (prod: the URL you open).
CLIENT_PORT	6856	Vite dev-server port (dev only: the URL you open with yarn dev).
CLAUDE_BIN	claude	The Claude Code binary to spawn. On Windows a bare name is resolved on PATH before it reaches the PTY layer (which matches file names exactly): to the .exe when there is one, otherwise to the .cmd shim an npm-global install leaves, run through cmd.exe.
CLAUDE_CWD	current dir	Working directory each claude PTY runs in; determines which project's sessions the sidebar lists. Via npx mulmoterminal@latest it defaults to the directory you ran the command from (override with --cwd <dir>, relative allowed); when the server is run directly it falls back to ~/mulmoclaude. A value read from .env must be an absolute path (~ is not expanded).
CLAUDE_PERMISSION_MODE	auto	Permission mode passed to each claude spawn.
MT_TITLE_MODEL	haiku	Model used for the cell header's AI title (a cheap/fast model summarizing the recent turns). Accepts a --model alias or a full model id.
CODEX_BIN	codex	The Codex CLI binary to spawn.
CODEX_MODEL	codex default	Model passed to Codex as --model (unset = Codex's own default).
CODEX_HOME	~/.codex	Codex home — where its session rollouts and MulmoTerminal-mirrored skills live.
ANTIGRAVITY_BIN	agy	The Antigravity CLI binary to spawn.
ANTIGRAVITY_MODEL	agy default	Model passed to Antigravity as --model (unset = agy's own default).
ANTIGRAVITY_HOME	~/.gemini/antigravity-cli	Antigravity home directory containing session brain storage.
MULMOTERMINAL_HOME	~/.mulmoterminal	Root for managed git worktrees.
CLAUDE_CONFIG_DIR	~	Claude Code's own config directory. .claude.json lives inside it, so relocating your Claude Code config moves that file too — MulmoTerminal reads it to tell whether the per-project GUI MCP server is registered (server/infra/gui-mcp-registration.ts). Leave it unset and ~/.claude.json is used.
MULMOCLAUDE_WORKSPACE_PATH	~/mulmoclaude	Where the managed MulmoClaude workspace lives. MulmoTerminal seeds presets/helps only into this directory, so launching in an arbitrary project never writes them there (server/backends/workspaceSetup.ts). Set it to the same value MulmoClaude uses.
MULMOTERMINAL_NO_SKILL_INSTALL	unset	Set to any value to skip installing the bundled skills (mulmoterminal-config and the -dirs / -theme / -header / -keys / -model / -notify / -bug-report / -decisions family) into ~/.claude/skills/ and the Codex skills root on startup.
GEMINI_IMAGE_MODEL	gemini-3.1-flash-image-preview	Model used for image generation (needs GEMINI_API_KEY). The default is a preview model Google schedules for retirement around mid-2026, so pin a stable one here (e.g. gemini-2.5-flash-image) rather than waiting for a code change.
WAIT_REAP_GRACE_MS	1800000	How long a waiting background session is kept before it's auto-reaped (0 or negative = never).

The Docker-sandbox variables (MULMOTERMINAL_SANDBOX, MULMOTERMINAL_SANDBOX_IMAGE, SANDBOX_MOUNT_CONFIGS, SANDBOX_SSH_AGENT_FORWARD) and the update-check opt-outs (MULMOTERMINAL_NO_UPDATE_CHECK, NO_UPDATE_NOTIFIER) are covered in Docker sandbox and Install & run.

Example .env (gitignored):

CLAUDE_CWD=/Users/you/my-project

UI settings (~/.mulmoterminal/config.json)

The Settings modal (⚙) persists per-user UI choices to ~/.mulmoterminal/config.json (read/written via GET/POST /api/config):

Open it from the ⚙ button in the toolbar. Pick a theme, set the terminal font size and scroll speed, set a custom attention sound, list the repos the cross-repo PRs & Issues view should aggregate, add launch commands for grid cells, and register your own MCP servers — no need to hand-edit the config file. Note that theme, font size and scroll speed are stored per browser (they're display preferences, so a phone and a desktop keep their own); the rest live in ~/.mulmoterminal/config.json and are shared by every client.

Field	Meaning
cwdPresets	Quick-pick directories offered when launching a terminal.
soundFile	Absolute path to a custom attention sound, the fallback for every kind. Empty/unset uses the built-in synthesized chime.
soundKinds	Which moments beep — see Notification sounds. Defaults to ["finished","waiting"]; the other kinds are opt-in.
sounds	Per-kind sound: { "waiting": "preset:coin" }. A preset:<id> reference or an absolute path; a kind with no entry falls back to soundFile.
prRepos	owner/repo entries whose open PRs/issues the cross-repo PRs & Issues view aggregates (via your gh login).
launchers	{ label, command } entries offered in a grid cell's launcher besides the agents — any interactive command. A plain shell needs no entry: the launch form's Shell toggle opens $SHELL unconfigured.
quickCommands	{ label, text, agents? } phrases the phone offers as chips on a session's terminal view. Tapping one puts text in the input box; it is not sent until you press send. agents ("claude" / "codex" / "shell") scopes a chip to session kinds — omit it to offer the chip everywhere. Empty by default.
userMcpServers	{ id, url } HTTP MCP servers merged into the single-view Claude session's --mcp-config (a localhost URL is reached over host.docker.internal in the Docker sandbox). Takes effect on the next session.
buttons	Header action buttons — see Header buttons. Omit to keep the defaults; set to replace them.
chips	Header info chips (dir / git / work / diff / ctx / usage / status / tools, or custom text). Omit to keep the default set; [] hides all built-ins. work shows which PR / issue the cell is on (#977 → #966) and clears itself when the PR merges — see the Configuration guide.
pushEnabled	true to send a Web Push to your registered devices. Off by default; only sends while the RemoteHost channel is connected (see below). The master switch — pushKinds picks which moments.
pushKinds	Which moments push: "finished" (a turn ended, ✅) and/or "waiting" (the agent stopped to ask — a permission prompt or a question, ❓, once per prompt). Omit to keep both; [] for none. A kind added in a later version stays off until you tick it.
worklogEnabled	true to run the built-in dev worklog batch (see below). Off by default (each run spawns an LLM session, so it costs tokens).
worklogIntervalHours	Worklog cadence in hours (default 6, clamped to 1–168).
terminalSubmit	Which bytes Claude reads as submit vs newline: "cr" (default — Enter submits, Shift+Enter makes a newline) or "esc-cr" (for a Claude Code rebound the other way). Applies to the keyboard and the phone remote-view submit, for Claude sessions only (shell/codex keep plain Enter). See the Configuration guide.
copyOnSelect	true puts a mouse selection on the clipboard the moment it settles, with no key pressed (the PuTTY / iTerm2 behaviour). Off by default — it changes the clipboard when you may only have meant to highlight something. No Settings UI: edit the file and reload the tab. Composes with the copy keymap action rather than replacing it. Over plain http:// the browser gives a page no clipboard access, so a fallback asks xterm to copy instead; see the Configuration guide.
decisionDigest	Keep a Markdown digest of the decisions this project's sessions asked for, refreshed at startup and every few hours, so an agent can read what has already been decided before asking something similar. Written to ~/.mulmoterminal/decisions/<project>.md (never into your repository) and served to agents by the bundled mulmoterminal-decisions skill. Off by default — it is a vision-stage idea, and it writes a file that would otherwise not exist. The digest holds dated facts, never inferred rules.
issueWorkComments	Let a cell comment on the issue it is working on: once when it starts, and again when its PR merges (closing the issue if GitHub has not already). The comment names the working directory it happened in — the folder name only, never the path — so a reader can tell which clone. Off by default; it writes to GitHub, often on somebody else's issue. Needs gh logged in. See the Configuration guide.
prWorkdirFooter	Ends a PR body with work in <clone> — the directory name of the clone the work happened in, so a PR says which of several side-by-side checkouts produced it. Applies to both paths that open PRs here: ⧉ Open PR appends it to the PR it creates, and every Claude session is told to end the bodies it writes with the same line (the name is resolved by the server, so a session inside a managed worktree still names the main checkout). On by default; set false to opt out — read per PR and per session spawn, so no restart is needed (there is no Settings control for it). Appending is idempotent: an existing PR never gets a second copy.
appendSystemPrompt	Whether a spawned Claude session is asked to end a reply with a closing summary — what was asked, what was achieved, what was not (see Closing summary). On by default; set false to opt out, and a directory's .mulmoterminal.json outranks this. Read per spawn, so no restart is needed (there is no Settings control for it), though a session already running keeps what it was launched with. true / false only.
fontFamily	The terminal font every session renders in — a CSS font-family stack, e.g. "'Cica', 'MS Gothic', monospace". No Settings UI: edit the file, then restart (this config is read once at startup). Unset uses the built-in stack (JetBrains Mono / Fira Code / Menlo / Consolas, then CJK faces for Japanese, Korean and Chinese). Unlike the per-browser font size, this is one value for the whole host — it names fonts, and which fonts exist is a property of the machine. A directory can override it. See the Configuration guide.

Every MulmoTerminal on the machine shares this one file, so an older build could save over a key a newer one wrote. It doesn't: a top-level key this version doesn't recognise is written back untouched, which is what makes running two versions side by side — or downgrading for a while — safe. A mistyped key survives on the same rule, which is deliberate: a line you can still see is easier to debug than one that silently vanished. See the Configuration guide.

Header buttons

Each terminal header shows configurable action buttons. Omitting buttons (globally or per-dir) keeps the built-in starter set: a file-path picker (📎), an OS file-manager reveal (📂), an in-app file explorer (📁), a new terminal here (🖥), this branch's PR (🔗, git repos, only when a PR exists), and open-on-GitHub (🌐, git repos). Setting buttons (at either level) replaces the whole default set with your list (it is not merged on top), so listing your own — even a shorter one — is how you drop, reorder, or swap them. A button has an id, label, and a run of "shell" (run a command), "input" (send text to the agent), or "open". An open button targets one of url / reveal (OS file manager) / files (in-app explorer) / view (a built-in overlay) / terminal (a dir → a new cell running $SHELL, opened next to the current one) / pr: true (open the current branch's PR — the button is hidden when there's no open PR) / pickFile: true (OS file dialog → insert the path). ${dir}, ${branch}, ${repo}, … substitute live context, and when (e.g. "isGitRepo") gates visibility. The /mulmoterminal-header skill writes a valid config interactively; per-dir buttons merge over the global ones by id, while chips replace the global list wholesale.

Notification sounds

Six moments can beep, each with its own sound and its own on/off switch. Running many agents at once is what turns notifications into noise, so only the first two are on by default — the rest are opt-in from Settings.

Kind	When	Default
finished	the turn ended and the output is unread	on
waiting	it stopped to ask — a permission prompt or a question	on
command-done	a Run cell's command exited 0	off
command-failed	a Run cell's command exited non-zero, or never started	off
session-exited	a session's terminal ended — including when you close the cell yourself	off
pr-ci-failed	a directory's PR went red. Only seen while the roster is on screen, since that is what polls the phase	off

A Run cell is the one-shot cell a script.json entry or a run:"shell" header button opens — not a shell launcher cell. A launcher runs an interactive shell that stays alive, so nothing marks where one command inside it ended; only the one-shot cell reports an exit code.

finished and waiting reach the phone too (pushKinds); the other four are seen only in the browser — a Run PTY never enters the session registry, and a PR phase is something the page polls — so Web Push cannot raise them.

What each one plays. The default chime is generated with the Web Audio API — no audio file is bundled, so the npm package stays light and has no media-licensing concerns. Beyond it there are two options:

• Presets — seven sounds hosted in the ownplate repo (MIT), referenced as preset:<id>: chime coin cheep door gong magic meow. The first play downloads one into ~/.mulmoterminal/sounds/; every later play reads that file, so a preset keeps working offline. A failed download is not remembered as one — you get the chime that time and the next play retries. That holds on both sides: the server caches no failure, and it answers 503 (not 404) for a preset it could not fetch, because the browser remembers a 404 for the life of the page and only retries a 5xx.
• Your own file — an absolute path, per kind in sounds or as the all-kind soundFile.

Resolution per kind, nearest first: the session directory's sounds[kind], its sound, your sounds[kind], your soundFile, then the chime. The server streams whichever applies at GET /api/sound?kind= / GET /api/dir-sound?cwd=&kind=, and the client falls back to the chime if it's missing or not audio.

Web Push on task finish. Enable pushEnabled in Settings to have the server send a push (title = the project dir, body = the last prompt) to your registered devices each time a background task finishes — the same signal as the attention chime, but for the panes you're not watching. Delivery is handled by the separate mulmoserver sendPush Cloud Function; MulmoTerminal only makes the call, and only while the RemoteHost channel is connected (its Google sign-in supplies the notification auth). With RemoteHost disconnected, or with no device registered, the toggle is a no-op.

Dev worklog (cross-clone). Set worklogEnabled: true in ~/.mulmoterminal/config.json (and restart — the scheduler reads its tasks at boot) to register a built-in scheduled task. Every worklogIntervalHours (default 6) it spawns a Claude session that reviews the work you did across all your saved working dirs (cwdPresets) since it last ran, and writes it up as a short manager-style report. Multiple clones/worktrees of the same repo (e.g. myapp, myapp2) are merged into one per-repository section, each covering what problem was addressed, what got solved, what's still in progress, and — mined from the transcripts — decisions that were only discussed and not built. The window is since the last run (tracked in config/scheduler/worklog-state.json), not a fixed 6 h, so a missed/slept run doesn't drop work. It reads and reconciles progress against vision.md / milestones.md (creating empty ones if absent) so a long-running goal isn't forgotten.

Output lands in the wiki: one weekly page per ISO week (data/wiki/pages/dev-log-YYYY-www.md — filenames are lowercase, or the wiki can't open them), each tagged worklog. To browse them, open the 作業ログ 一覧 hub page (worklog), which links every week, or click the #worklog tag in the wiki index.

Off by default because each run costs tokens — watch the cost readout and tune the cadence. Run it on a single "hub" instance; running it in several instances sharing one workspace double-fires it. The batch treats everything it reads (transcripts, git, wiki) as untrusted data and only writes the worklog / hub / vision / milestones pages.

Per-directory settings (<project>/.mulmoterminal.json)

Drop a .mulmoterminal.json in a project directory to give terminals opened in that directory their own look and sound. It applies per terminal (per grid cell) — the rest of the app keeps your chosen theme — and a directory's theme overrides your manual theme pick for that terminal only. Every field is optional; a missing or malformed file is ignored.

{
  "name": "PROD · payments",            // badge shown on this directory's terminals
  "badgeColor": "#cf222e",              // badge color (hex #rrggbb)
  "headerColor": "#190a23",             // cell header background (hex #rrggbb)
  "headerTextColor": "#ffffff",         // cell header text color (hex #rrggbb)
  "cellColor": "#101014",               // cell body background (hex #rrggbb)
  "cellBorderColor": "#2a2a4e",         // cell border color (hex #rrggbb)
  "dotColor": "#00e676",                // idle status dot (hex #rrggbb)
  "buttonColor": "#c7cdf0",             // header icon buttons (hex #rrggbb)
  "theme": "nord",                      // terminal palette: midnight | nord | daylight | solarized
  "colors": { "background": "#190a23", "cursor": "#ff2e63" }, // per-key palette overrides
  "fontSize": 16,                       // terminal font size in px (8–32); overrides Settings
  "fontFamily": "'Cica', monospace",    // terminal font stack; overrides the global config
  "orderPriority": 10,                  // rank in the grid's "priority" order and the launcher chips (lowest first)
  "sound": "./.mulmoterminal/alert.mp3", // attention sound, RELATIVE to this directory
  "sounds": { "command-failed": "preset:gong" }, // per-notification-kind override
  "appendSystemPrompt": false           // no closing summary here; omit to follow the global setting
}

As cells pile up it gets hard to tell which project is which. Give each repo a name badge and its own colors in .mulmoterminal.json and they're unmistakable — headerColor/badgeColor tint the frame, while colors reaches all the way into the terminal's own background and text. (The example above dresses four repos in Mondrian / van Gogh / Picasso / Matisse palettes.)

Field	Meaning
name	Label shown as a badge in the terminal/cell header.
badgeColor	Badge background color (#rrggbb); text auto-contrasts.
headerColor	Header background color (#rrggbb) — the grid cell's header row and the terminal's own header row (grid row 2 + single view). While a terminal is working/blocked the status tint still shows; the custom color applies when idle.
headerTextColor	Header text color (#rrggbb) — the dir path, title, and prompt.
cellColor	Cell body background color (#rrggbb) — the frame around the terminal.
cellBorderColor	Cell border color (#rrggbb). The status frame (working/blocked) still overrides it while active.
dotColor	Idle status-dot color (#rrggbb). The working/waiting colors are unchanged so the activity signal stays intact.
buttonColor	Header icon button color (#rrggbb) — expand / close / attach / folder / etc., across both header rows.
theme	xterm palette for terminals in this directory (one of the built-in theme ids).
colors	Per-key xterm palette overrides applied on top of theme (or the app theme when theme is unset). Keys are xterm ITheme names (background, foreground, cursor, selectionBackground, the 16 ANSI colors, …); values are hex (#rgb / #rrggbb / #rrggbbaa). Unknown keys / bad values are dropped.
fontSize	Terminal font size in px for this directory (8–32), overriding the Settings value. A size outside the range is clamped; a non-number is ignored. Changing it re-fits the terminal, so the PTY learns the new width — unlike browser zoom, which leaves the two disagreeing.
orderPriority	This directory's rank in the grid's priority ordering — the third mode on the toolbar's ordering button, next to auto (attention-first) and manual (the move buttons). Any integer, lowest first; negatives are allowed. Directories that set nothing sort last, keeping their existing order, so adding the key to one project doesn't shuffle the rest. The grid reads it in priority mode only; the launcher's directory chips always sort by it, so a project sits in the same place on both.
fontFamily	CSS font-family stack for this directory's terminals, overriding the global fontFamily. Use the names as your OS lists them ("'Cica', 'MS Gothic', monospace"). An unusable stack is ignored whole rather than half-applied; monospace is appended if you name no generic family. Prefer fonts whose fullwidth glyphs are exactly twice the Latin width, or box-drawing frames tear.
sound	Attention sound for this directory's sessions, a path relative to the directory (served at GET /api/dir-sound). The fallback for every kind.
sounds	Per-kind override of sound: { "command-failed": "preset:gong" }. Each value is a preset:<id> or a directory-relative path, under the same confinement.
appendSystemPrompt	Whether this directory's Claude sessions are asked to end a reply with a closing summary (see Closing summary). Omit to follow the global appendSystemPrompt, which is on; true / false here outranks it. Read per spawn, so a new session in this directory picks up an edit without a restart.
addDirs	Extra directories this project's Claude sessions may read and edit — the terminal-side equivalent of opening several folders in one VS Code workspace, via Claude Code's --add-dir. Relative entries resolve against this file's directory ("../shared-lib"), a path that doesn't exist is dropped, max 16. In the Docker sandbox each one is bind-mounted too, so the grant is real inside the container — which widens the sandbox on purpose. Claude only: codex has no equivalent flag and ignores the key.

Security. sound and every sounds entry are directory-relative paths only — absolute paths and any ../ that escapes the directory are rejected, and the path is never taken from the HTTP request, so an opened project can't point the player at arbitrary files. When changes take effect. A write made through Claude's tools — which includes the mulmoterminal-dirs skill — applies live: the tool hook that reports the write doubles as the reload signal, so colors, palette, font size and grid order update without reopening anything. There is no filesystem watcher, so an edit made outside a session (your own editor) is picked up when the terminal is next opened.

Checking what took effect. Settings → Directory settings lists your recent directories and expands each one to the values in force, with a swatch per color and the path of the file they came from. It also names the keys it dropped (a color that isn't #rrggbb, a size out of range) and the keys it doesn't read at all (badgeColour, a global-only setting) — which is what tells "I never set that" apart from "I set it and it didn't take".

---

Running

yarn install            # postinstall fixes node-pty prebuilt binary perms

yarn dev                # backend (:34567) + Vite UI (:6856), concurrently — open http://localhost:6856
# or individually:
yarn dev:server         # backend only  (node --import tsx --env-file-if-exists=.env server/index.ts)
yarn dev:client         # Vite dev server only

yarn build              # type-check (vue-tsc) + vite build -> dist/
yarn typecheck:server   # type-check the server (tsconfig.server.json)
yarn typecheck:test     # type-check the specs (tsconfig.test*.json)
yarn server             # run backend; serves dist/ + the APIs on :34567
yarn test               # vitest run

The backend is TypeScript run directly via tsx (no build step); server/ is type-checked separately through tsconfig.server.json (strict), kept out of the main build so the two type-check independently.

Specs sit outside both of those projects, and vitest strips types rather than checking them — so yarn typecheck:test is what keeps them honest. It mirrors the same split: tsconfig.test.json (client specs, DOM + .vue) and tsconfig.test-server.json (server specs, node). CI runs it alongside the other two.

In dev, open the Vite URL; its proxy forwards /ws, /ws/pubsub, and /api to :34567. In production, run yarn build then yarn server and open http://localhost:34567.

---

Scripts (Run menu)

An empty grid cell's launcher sets the Working directory by typing, by a preset chip, or with the 📁 folder button (a native OS folder dialog). It also offers a run a script row that launches project scripts (a dev server, tests, a build, …) in that cell, in the directory the cell is pointed at — so a whole workflow lives in one window alongside the Claude sessions. Scripts are per-directory: the cell reads the script.json of whatever directory you select, so different cells can offer different projects' scripts.

The same launcher also has an or launch row for your configured launch commands — any interactive command — set in Settings (⚙) → Launch commands as { label, command } (e.g. htop → htop, Codex → codex). A plain shell needs no entry here: the launch form's Shell toggle already opens $SHELL. Unlike a one-shot script, a launcher runs as a persistent terminal in the cell's directory: it survives grid page switches and reconnects, and its dot shows running vs. exited (it has no Claude hooks, so no blocked/done states).

Every running terminal's header also has a ▶ Run ▾ dropdown (next to the connection status), in both the single view and each grid cell — but only when the open project has scripts (no script.json, no button). It lists the open project's script.json — the directory that terminal runs in — and launches the picked script in a spare grid cell (reusing an open launcher, else a new one), switching to the grid from the single view so you can watch it. So you can start a dev server or tests for the project you're working in without disturbing the session that's running.

The list is populated from a script.json at the chosen directory's root. It's optional; a directory without one simply shows no scripts.

// <dir>/script.json
{
  "scripts": [
    { "label": "Dev server", "command": "yarn dev" },
    { "label": "Unit tests", "command": "yarn test" },
    { "label": "Build", "command": "yarn build" },
    // optional per-script working dir (relative to this file, or absolute):
    { "label": "Sub server", "command": "yarn serve", "cwd": "packages/server" }
  ]
}

Field	Required	Meaning
label	yes	What the launcher shows.
command	yes	Shell command, run via the login shell ($SHELL -lc "<command>").
cwd	no	Working dir, relative to script.json or absolute. Defaults to the cell's directory.

A command terminal is not a Claude session: it has no session id, no hooks, no transcript, and isn't persisted — it's ephemeral, so a page reload drops it and closing the cell (or reloading) kills the process. When the command exits, the cell offers a ↻ re-run. The browser only ever sends the script's index + its directory; the server reads that directory's script.json and resolves the command, so the file is the allowlist of what can run.

Each command cell also has a ✦ Summarize button: click it to send the cell's captured output to claude -p (headless) and get a short Errors / Warnings / likely cause / suggested fix note in a panel — handy when a build or install buries the one failing line in thousands. It's manual (never auto-runs) and analyzes the last 32 KB of output. See POST /api/command/summarize.

---

Skills (Skill menu)

Next to the ▶ Run ▾ dropdown, every running terminal's header has a ⚡ Skill ▾ dropdown — in both the single view and each grid cell, and only when the open project has skills (nothing discovered, no button). It lists the Claude skills discoverable for that terminal's directory — both project scope (<dir>/.claude/skills) and user scope (~/.claude/skills), the same skills Claude sees — and, on pick, runs the skill in that session: it types the skill's invocation into the terminal and submits it (for Claude, its /<slug> command; for Codex, which has no slash command, a plain Use the "<slug>" skill. instruction). Unlike ▶ Run — which launches a script.json shell command in a spare cell — a skill runs in the session you picked it from, continuing that conversation.

Ordering: working-dir (project) skills come first, then user-scope ones, alphabetical within each group; a project skill of the same slug shadows the user one.

Filtering: add a skills array to the directory's .mulmoterminal.json to narrow the menu — an allowlist of slugs that also sets the order (only those show, in that order). Omit it to show everything.

// <dir>/.mulmoterminal.json
{ "skills": ["review-diff", "commit-msg"] }

Each menu item shows the skill's id, with its SKILL.md description as the hover tooltip. A directory (or workspace) without any .claude/skills simply shows no button. Skills are discovered read-only; the menu never creates or edits them.

---

Files view (browse & edit)

A terminal header can carry a 📁 Files button — add it as a header button ("open": { "files": "${dir}" }) — that opens a full-screen file explorer rooted at that terminal's project directory — so after Claude says "wrote foo.md" you can jump straight there to read or edit it. The left pane is a lazy-loaded directory tree; clicking a file opens it in a CodeMirror editor (Markdown / JS-TS / JSON highlighting, everything else as plain text). Markdown files get a Preview toggle that renders via the server's sandboxed …/md HTML. Save (or ⌘/Ctrl-S) writes back.

Beside an enlarged terminal, not only full-screen. Expand a grid cell (⤢) and its header gains a folder toggle that splits the enlarged area in two: terminal on the left, the same explorer + editor on the right, rooted at that cell's directory. Drag the divider (or focus it and use ←/→, Home, End) to resize — the terminal keeps a floor, so a squeeze shrinks the pane rather than reflowing xterm into garbage. It works in both zoomed layouts (cockpit roster and thumbnail filmstrip), the pane re-roots as you walk the zoom between terminals, and whether it's open plus how wide it is are remembered per browser.

The toggle is not the only way in: while a cell is enlarged, clicking a file path the agent printed opens it here too, rather than in a new tab or full-screen — see Clicking a file path.

All reads and writes go through GET/PUT /api/files/browse/*?cwd=&path=, and every path is contained within the project root (server-side) — ../absolute escapes are rejected for reads and writes alike, so editing can't reach outside the directory the terminal is pointed at. A save sends the version the file had when it was opened, so it is refused (409) rather than silently overwriting an agent that edited the same file meanwhile; the editor then offers to reload or to overwrite deliberately.

You usually hear about it before that. An open file that changes on disk is picked up from Claude's own write hook (immediately) and from a 30-second version check (which catches Codex, git, builds and other editors too). A clean buffer just takes the new content — the pane reads as a live view — while a dirty one raises the same banner rather than choosing for you.

Leaving an open file saves it — switching files, moving the enlargement to another terminal, closing the pane, navigating away. No dialog interrupts you mid-flow, because opening a file, and replacing one, keep a copy under ~/.mulmoterminal/backups/ — three generations per file, outside the project so they never reach git status or the agent's view of its own repo. A parting save that loses the version race banks your version there instead of overwriting the other writer. Re-opening unchanged content doesn't rotate one in, and a backup that can't be written never blocks the read or the save it was taken for.

---

Git worktrees & pull requests

When a terminal's directory is a git repo, its header shows a branch chip (⎇ <branch> with dirty / ahead / behind counts), fed by GET /api/git-status (polled while the view is visible). A GitHub menu links straight to the repo, its issues, and its pull requests.

Worktree isolation. A grid cell's launch form offers ＋ New worktree: name a task and the cell launches its agent inside a fresh git worktree on a new agent/<slug> branch — a separate working tree that shares the repo's .git, so several agents can work the same repo without colliding. Worktrees live under ~/.mulmoterminal/worktrees/ (override with MULMOTERMINAL_HOME), and existing ones are listed for reuse.

Every empty grid cell shows this launch form: toggle Claude / Codex / Antigravity / Shell, type a working directory (frequent ones autocomplete from your presets), or — in a git repo — name a task under OR ISOLATE IN A WORKTREE and hit ＋ New worktree to start the agent on its own isolated branch. Shell runs your OS default shell there instead of an agent; OR LAUNCH runs one of your configured launch commands.

A worktree cell's header carries a diff badge (+<commits> ●<dirty>); click it for a Changes vs <base> panel (file list + patch) with actions:

• ✓ Commit — hands the cell's own session a canned commit prompt.
• ⬆ Push — git push -u origin <branch> (POST /api/worktrees/push).
• ⧉ Open PR — pushes, then gh pr create … --fill; if gh is missing or unauthed it falls back to opening the GitHub compare URL (POST /api/worktrees/pr).

Closing a worktree cell asks whether to keep the worktree or discard & remove it (a dirty worktree is never removed unless you confirm).

PRs & Issues (cross-repo). The toolbar's Pull requests button opens a full-screen view that aggregates open PRs and issues across the repos listed in Settings → Pull request repos (prRepos, owner/repo entries) via your server-side gh login. PRs show a CI-rollup / review-decision / draft badge; each repo lists its latest open issues. Rows are real links, per-repo errors don't sink the view, and the two lists load independently. Backed by GET /api/prs and GET /api/issues.

---

Cost & token usage

Each grid cell's header shows two badges for its session, refreshed when a turn finishes (from GET /api/session/:id):

Both badges, live on a real Claude session: Opus · ctx 5% (model family + how full its context window is) and ⇡427k ⇣1.8k (cumulative input / output tokens for the session). They sit in the header's first row alongside the status dot, directory, and git chip (⎇ main ●2), with what the agent is doing to the right; the icon buttons and the timeline (🕘) of tool calls are on the second row.

• Context badge — e.g. Opus · ctx 35%: the model family plus how full its context window is (the last turn's input + cache tokens ÷ the model's window — 1M for current-gen Opus / Sonnet / Fable / Mythos, 200k otherwise). A session running on a provider model shows that model's name and its published window (Kimi K2.7 Code · ctx 12%); a model in neither list keeps the label and hides the %, since the window is never guessed. A reading past 100% shows ctx ? instead of the number: the window is a hard cap, so an impossible percentage means the built-in window table is out of date for that model rather than that the session is over-full.
• Token badge — ⇡<in> ⇣<out>: cumulative input (fresh + cache-read + cache-creation) and output tokens for the session, k/M-formatted, with a full breakdown in the tooltip.

The Settings modal (⚙) shows an estimated $ cost — Session / Today / Month — from GET /api/cost, using a built-in public per-model price table (cache reads billed at 0.1×, cache writes at 1.25× input). It's an estimate: real billing differs, flat-plan (Max) usage isn't reflected, and turns on unpriced models are flagged and excluded.

A separate, full double-entry accounting book (the account_balance toolbar button → /accounting) is provided by the bundled @mulmoclaude/accounting-plugin and stores its books under <workspace>/data/accounting. It's a bookkeeping app — unrelated to the LLM cost estimate above — and is also exposed to Claude as the manageAccounting GUI tool.

---

Wiki, Collections & the GUI panel

MulmoTerminal is also a live view over the shared workspace (CLAUDE_CWD, default ~/mulmoclaude) that agents author into — never a snapshot, so it re-reads on entry.

GUI panel. Beside the terminal, a GUI panel ("Canvas") renders the rich results of GUI-protocol tools the agent calls — documents (presentDocument), forms (presentForm), generated images, charts, HTML, and collection cards. Each result is drawn by its plugin's own Vue view inside a Shadow-DOM PluginFrame (so a plugin's bundled CSS can't leak), mirrors the active session, and replays history on re-select. Plugins reach the agent over an in-process MCP server served per session at POST /api/mcp/:sessionId (server name mulmoterminal-gui) — which works from the host or the Docker sandbox (over host.docker.internal). Which plugins load is gated by plugins/plugins.json; the shipped set includes markdown, form, image generation (needs GEMINI_API_KEY), chart, HTML, collection, and mulmoscript (MulmoCast video/slides/PDF playback) views. You can also merge your own HTTP MCP servers into the single-view session via Settings → userMcpServers.

Wiki. The toolbar Wiki button opens a read-only browser over <workspace>/data/wiki/ — an index (tag-filterable page catalog), rendered pages with [[wiki links]] and backlinks, a graph view (pages ranked by references), and a lint report (orphans / broken links / tag drift) whose [[links]] are clickable too. Read-only endpoints: GET /api/wiki, /api/wiki/graph, /api/wiki/lint.

Collections. The toolbar Collections button browses the workspace's collection "cards" (@mulmoclaude/collection-plugin). Running a collection action fetches a seed prompt and spawns a fresh agent session for it — the Launch with Claude / Codex toggle decides which agent (and whether the seed auto-runs or drops in as an editable draft). Favorited collections get their own toolbar buttons.

---

More features

• Grid of parallel sessions — the ＋ Terminal / grid view runs many sessions at once, auto-sizing by count across pages. Cell borders signal state at a glance — working (pulsing blue), blocked (amber — needs a permission / answer), done (blue — finished, output unreviewed), and idle — and the toolbar shows a tally across all pages so you notice an off-screen cell that needs you.
• Zoom & filmstrip — a cell's ⤢ enlarges one agent while the rest shrink to thumbnails in a bottom filmstrip; click a thumbnail to switch, ⤡ to return to the grid — so you can flip between "see everything" and "focus on one" in a click. While zoomed, keys you bind walk the enlargement along the on-screen order without reaching for the mouse — opt-in, nothing is bound by default, since any bound key is taken from the terminal underneath. Add a keymap to ~/.mulmoterminal/config.json; see the guide for the syntax, the action list, and combinations a browser can never bind.

• Timeline (🕘) — a read-only per-session activity timeline (tools run, newest first), from GET /api/transcript/timeline.
• Bring another cell's turn here (💬) — pick another terminal in the grid and its last completed turn is pasted into this cell's input box, so you can have Claude and Codex look at each other's work (or pull in a session running in a different repo). The excerpt comes from the agent's own log, not the screen buffer, so it carries no ANSI debris and nothing lost to scrollback. It is pasted, never sent — you read what arrived and press Enter, in the cell you were already in. A turn still running isn't available yet (Codex writes its rollout only once the turn ends).
• Tools pane — the available GUI tools plus a live tool-call history for the active session.
• Notifications (🔔) — a toolbar bell with an unread badge and a dropdown of active notifications; click a row to jump to its session.
• Star MulmoTerminal — a star button in the grid toolbar that stars the project on GitHub through your own gh login, in one click. It is a one-time ask: once the repo is starred the button is gone for good and stops calling the server at all. It shows only when gh can answer — with no gh, no login, or no network, one click couldn't star anything, so nothing is shown and nothing is recorded. Set gh up later and the button appears by itself.
• Voice input — dictate a prompt via on-device Whisper (POST /api/transcribe, macOS only; the model downloads on first use). Settings picks the language you dictate in (per browser): your browser's, whisper's own per-clip detection, or a fixed one. Worth setting — speech in a language the mic is not expecting comes back translated into the one it is, so an English browser silently turned Japanese dictation into English.
• Remote host — link MulmoTerminal to the companion phone client (Google sign-in) to watch and start sessions from your phone.
• Themes — four terminal palettes (midnight / nord / daylight / solarized), your pick remembered; a project's .mulmoterminal.json can override per directory.
• Editing niceties — Shift+Enter inserts a newline in the prompt, and on macOS Option is treated as Meta so Claude's Alt-key bindings work. If your Claude Code is rebound so Enter and Shift+Enter behave backwards, flip them with terminalSubmit.
• Scroll speed — one wheel notch or trackpad swipe moves the terminal the same distance whether you're reading a shell's scrollback or a full-screen app like Claude Code. If a two-finger scroll on a Mac trackpad flies past what you were reading, turn terminal scroll speed down in Settings (0.25×–3×, per browser — it's a property of the pointing device).
• No accidental page zoom — Ctrl+wheel and a trackpad pinch would rescale the whole page and drag the layout and the terminal's fit along with it, so both are ignored. Keyboard zoom (Cmd/Ctrl + / -) still works when you mean it, and a phone's finger pinch is untouched. To make terminal text bigger for real, use the font size in Settings (or a directory's fontSize) — that re-fits the PTY instead of leaving it disagreeing.

---

Server API specification

Base URL: http://localhost:$PORT (default http://localhost:34567).

HTTP: GET /api/sessions

Lists the most-recent chat sessions for the current project (CLAUDE_CWD), newest first, including freshly-created sessions that aren't yet written to disk.

Response 200 application/json

{
  "cwd": "/Users/you/my-project",
  "sessions": [
    {
      "id": "d16f43f3-ef63-4a5e-b273-debaccb3522a", // session UUID (= .jsonl basename)
      "title": "Review available skills list",        // see "Session discovery & titles"
      "mtime": 1781471064511.22,                       // last-modified, ms epoch (sort key)
      "working": false,                                // Claude is mid-turn (blue dot)
      "waiting": false                                 // needs attention (bold)
    }
    // ...
  ]
}

• Sessions are read from ~/.claude/projects/<encoded CLAUDE_CWD>/*.jsonl and merged with in-memory sessions started this run but not yet persisted (those have title: "New session" and mtime = creation time).
• Sorted by mtime descending and capped at the 50 most recent. Files are ranked by a cheap stat-only pass; only the top 50 are read and parsed for titles, so the endpoint stays cheap regardless of how many sessions exist.
• 500 { "error": string } on an unexpected filesystem error. A missing project directory is not an error — it yields an empty sessions array.

Directories that cannot be used

Every ?cwd= — on the terminal sockets and on the read routes alike — names the directory the request is about. When one is named and cannot be used, the server says so instead of quietly answering about the default workspace (#1151):

Where	What happens
/ws, /ws/codex, /ws/antigravity, /ws/launch, /ws/run	The socket is closed with { type: "error", message }, which the terminal shows as a red banner and does not retry.
A session that is still running (?session= names a live PTY or a surviving tmux session)	Attaches anyway, with a warning in the server log. Moving or renaming a directory must not shut you out of an agent that is still working in it — and the cwd reported back comes from the running PTY, not from the request.
GET /api/scripts, /api/skills, /api/dir-config, /api/dir-sound, /api/git-status, /api/pr-phase, /api/header, /api/sessions, /api/codex/sessions, /api/session/:id, /api/transcript/*, /api/cost	404 { error, cwd } — a directory that is not there.
A ?cwd= that cannot name a directory at all (relative, or repeated as ?cwd=a&cwd=b)	400 { error, cwd }.

A request that names no directory is unaffected: CLAUDE_CWD is then the answer it asked for. The wording of the refusal is the one a refused spawn already uses, so the same condition reads the same whether it is caught here or by ptySpawn itself.

HTTP: GET /api/scripts

The runnable entries from <cwd>/script.json for a cell's chosen directory (?cwd=<dir>, or CLAUDE_CWD when none is named); see Scripts (Run menu). The resolved cwd is echoed back, and each entry carries its index (the position the client sends back to /ws/run). A ?cwd= that names a directory the server cannot enter is answered 404 { error, cwd } rather than with the default workspace's scripts — see Directories that cannot be used.

// GET /api/scripts?cwd=/Users/me/proj
{
  "cwd": "/Users/me/proj",
  "scripts": [
    { "index": 0, "label": "Dev server", "command": "yarn dev" },
    { "index": 1, "label": "Sub server", "command": "yarn serve", "cwd": "packages/server" }
  ]
}

A missing or invalid script.json is not an error — it yields an empty scripts array.

HTTP: GET /api/skills

The Claude skills discoverable for a terminal's chosen directory (?cwd=<dir>, or CLAUDE_CWD when none is named) — project scope (<cwd>/.claude/skills) plus user scope (~/.claude/skills), deduped by slug (project shadows user), working-dir skills first; see Skills (Skill menu). A skills allowlist in that directory's .mulmoterminal.json narrows and reorders the result; absent → all. The resolved cwd is echoed back. Each entry carries its slug (the skill invoked as /<slug>) and the SKILL.md description (the menu tooltip).

// GET /api/skills?cwd=/Users/me/proj
{
  "cwd": "/Users/me/proj",
  "skills": [
    { "slug": "commit", "description": "Write a commit message" },
    { "slug": "review", "description": "Review the current diff" }
  ]
}

A directory without any discoverable skills is not an error — it yields an empty skills array.

HTTP: POST /api/command/summarize

Runs claude -p headless over a command cell's captured terminal output and returns a short summary (Errors / Warnings / likely cause / suggested fix). Backs the ✦ Summarize button on a Run cell (see Scripts (Run menu)). The browser sends the cell's xterm buffer as log; the server truncates it to the last 32 KB (the tail, where errors + the exit line live), runs the CLI with the log piped on stdin (argv — no shell), and returns its answer. Same-origin guarded.

Request application/json:

{ "log": "npm ERR! cannot find module 'foo'\n..." }

Response 200 application/json:

{
  "summary": "Errors: cannot find module 'foo'\nSuggested fix: run `yarn add foo`",
  "truncated": false // true when the log exceeded 32 KB and only the tail was analyzed
}

Empty output returns a { summary } note rather than calling the CLI. Errors: 400 (missing log), 403 (disallowed origin), 502 (the claude run failed).

HTTP: POST /api/hook

Internal endpoint. Claude hooks (injected per session — see Claude hook injection) POST their event payload here. You normally don't call this yourself.

Request application/json — the Claude hook payload; only these fields are used:

{
  "session_id": "d16f43f3-...",        // the session the event is for
  "hook_event_name": "UserPromptSubmit" // "UserPromptSubmit" | "Stop" | "Notification"
}

Effect (see Session model):

hook_event_name	Effect
UserPromptSubmit	working = true for the session.
Stop	working = false; if the session is backgrounded, also waiting = true.
Notification	If the session is backgrounded, waiting = true.

Any resulting state change is published on the sessions pub/sub channel.

Response 200 application/json: { "ok": true } (always, even for unknown events).

More HTTP endpoints

The endpoints above are the core; the server exposes many more (all under http://localhost:$PORT; query params shown where relevant). Mutating endpoints are same-origin-guarded.

Sessions & agents

Endpoint	Purpose
GET /api/session/:id?cwd=	One session's summary — cumulative usage and context (model + last-turn context tokens). Backs the cell token & ctx% badges.
GET /api/codex/sessions?cwd=	Codex sessions for the project (from ~/.codex rollouts), newest first.
GET /api/cost?cwd=&session=	Estimated $ cost — session / today / month.
GET /api/transcript/timeline?session=&cwd=	Per-session activity timeline (tools run).
GET /api/transcript/last-turn?session=&cwd=&agent=	A session's last completed exchange (prompt, reply) plus the text to paste into another terminal. agent=codex reads the codex rollout instead of the Claude transcript.
GET /api/decisions?cwd=&limit=	The decisions a human was asked to make in this project, newest first — each question with the options it offered, their descriptions, and the answer. answerKind says whether the answer was one of the options, text the user wrote instead (the question was wrong), or never given. Read out of Claude's own transcripts; writes nothing. scanned reports how many transcripts were read (the scan is capped) and unreadable how many could not be, so a partial answer is visible rather than implied. A cwd that is not an existing directory answers an empty response rather than falling back to the default workspace.
GET /api/decisions/digest?cwd=	The same decisions as Markdown, for an agent to read ({ enabled, markdown }). enabled: false means the decisionDigest setting is off — a different answer from an empty digest, so a reader can tell "switched off" from "nothing decided here".

Git & worktrees

Endpoint	Purpose
GET /api/git-status?cwd=	{ repo, branch, detached, dirty, ahead, behind, upstream }.
POST /api/git-remote	The dir's GitHub repo URL (for the header GitHub menu).
GET /api/worktrees?cwd= · GET /api/worktrees/diff?cwd=	List managed worktrees / diff one vs its base.
POST /api/worktrees/create · /remove · /push · /pr	Create on agent/<slug>, remove (managed root only), push, open a PR (gh, else compare URL).
GET /api/prs · GET /api/issues	Open PRs / issues across the configured prRepos (via gh).
GET /api/github/star · POST /api/github/star	Whether you have starred MulmoTerminal, and star it (via gh). starred: null means gh could not answer, and hides the button.

Workspace views

Endpoint	Purpose
GET /api/wiki (?slug=) · /api/wiki/graph · /api/wiki/lint	Read-only wiki index / page / graph / lint.
GET /api/collections/… · /api/feeds · GET|PUT /api/shortcuts	Collections browser, feeds, favorites (see docs/collection-plugin-integration.md).
GET /api/files/browse/{list,text,version,md} · PUT /api/files/browse/{write,backup}	File tree / read / Markdown-render / write (contained within the project root). text answers { text, version }; write takes { text, baseVersion } (null = expecting to create it) and answers 409 with the version now on disk if the file changed since — so a save can't silently overwrite the agent that edits the same files. version answers that token alone, for the editor's periodic check. backup banks a buffer the editor is about to discard.
GET /api/files/raw?path=	Raw asset bytes (workspace-rooted).

GUI panel / plugins / MCP

Endpoint	Purpose
POST /api/mcp/:sessionId	Per-session GUI MCP server (Streamable HTTP; GET/DELETE → 405).
POST /api/plugin/:toolName	GUI-plugin dispatch (incl. spawnBackgroundChat, manageAccounting, presentHtml).
GET /api/agent/toolResults/:id · POST /api/agent/toolResult	GUI-panel result history / persist.
GET /api/tools · GET /api/tool-calls/:id	Available tools / tool-call history.
POST /api/accounting	Double-entry accounting (bundled plugin).

Config, sound & misc

Endpoint	Purpose
GET|POST /api/config	User UI config (cwdPresets, soundFile, soundKinds, sounds, prRepos, launchers, quickCommands, userMcpServers, providers).
GET /api/sound?kind= · /api/dir-sound?cwd=&kind= · /api/sound-preset/:id · /api/dir-config?cwd=	Custom / per-directory / preset attention sound + per-dir config. kind selects a config entry, never a path.
GET /api/dir-config-detail?cwd=	The same per-dir config, plus the settings a running terminal doesn't need (provider, model, skills, addDirs, header button/chip labels), plus which keys the file set and how each fared (applied / dropped in validation / not a setting at all). Read-only; backs the Settings modal's Directory settings preview. Unlike the other ?cwd= routes this one does not fall back to the default workspace — it reports on the directory it was asked about, so a path that no longer exists comes back as exists:false. Sound paths and button commands stay server-side.
GET /api/launch-options	The Anthropic-compatible backends this server can reach, each with its models and — when it can't — the reason. Reports the name of the env var a key is read from, never the key.
GET /api/notifications(/history) · POST /api/notifications/:id/clear	Notification feed.
POST /api/transcribe(/model…)	Voice-input transcription (Whisper, macOS).
POST /api/translation	Runtime UI-string translation.
GET /api/remote-host/status · POST /api/remote-host/{connect,disconnect}	Companion phone-client link. Each response carries the command channel's health (online / reconnecting / offline, plus the last listener error), so the toolbar shows a dropped channel instead of the last state it happened to fetch.
POST /api/open-dir · POST /api/pick-file	Reveal a dir in Finder/Explorer; OS file-picker → path ({ directory: true } opens the folder picker — used by the launcher's Working-directory 📁 button).
POST /api/session/:id/drop	A dropped file whose path the browser withheld. Raw bytes, not JSON, under the file's own content type (base64 in JSON would cap real files near 18 MB, and a dropped .json would be parsed as a document); the original name rides percent-encoded in x-drop-filename and is used for its suffix only. Answers { path } — absolute, inside the private per-session directory the session was granted at launch. 110 MiB cap; 404 for a session this server isn't running.

The phone itself uses none of these routes — it reaches the host over Firestore command docs, not HTTP. Every command it can send, and the shapes it gets back, are in docs/remote-host-protocol.md.

WebSocket: /ws (terminal)

A raw WebSocket carrying the terminal stream for one session. One PTY per connection (or reattach to an existing background PTY).

Connect

• ws://host/ws — start a new session (server generates a UUID and spawns claude --session-id <uuid> --settings <hooks>).
• ws://host/ws?session=<id> — resume/reattach a session. If a live background PTY exists for <id>, the socket reattaches to it (and its recent output buffer is replayed); otherwise the server spawns claude --resume <id> --settings <hooks>.

Server → client (JSON text frames):

Message	Meaning
{ "type": "session", "id": string }	Sent immediately on connect — the session id this socket is bound to (lets the client learn a new session's generated id).
{ "type": "output", "data": string }	PTY output to write to the terminal. On reattach, the first output frame is the replayed tail buffer (≤ 64 KB).
{ "type": "exit", "exitCode": number, "signal": number }	The claude process exited; the socket then closes.

Client → server (JSON text frames):

Message	Meaning
{ "type": "input", "data": string }	Keystrokes / bytes to write to the PTY.
{ "type": "resize", "cols": number, "rows": number }	Resize the PTY.

A non-JSON frame is written to the PTY verbatim (fallback).

Disconnect — when the socket closes, if Claude is still working the PTY is kept alive in the background; otherwise it's killed. See Session lifecycle.

More WebSocket endpoints

Two more raw WebSockets share the /ws frame format (output / input / resize / exit):

• /ws/codex?session=<id>&cwd=<dir>&gui=<0|1> — a Codex agent PTY (see Agents: Claude & Codex). Like /ws it sends a session frame with the id and reattaches to a live or tmux-backed session on resume. gui=0 (grid cells) omits the GUI MCP and keeps the session out of the sidebar.
• /ws/launch?session=<id>&cwd=<dir>&launcher=<index> — a launch command PTY (a plain shell, codex, or any command configured in Settings → Launch commands). Unlike a Run-menu script it's persistent and reattachable (survives page switches / reconnects), but it has no Claude hooks, so its dot only shows running vs. exited.

WebSocket: /ws/run (command terminal)

A raw WebSocket carrying a one-off Run-menu command (see Scripts (Run menu)) — a plain shell PTY, not a Claude session, so there's no session message, no hooks, and no reattach.

Connect

• ws://host/ws/run?index=<n>&cwd=<dir> — run the script at position <n> in <dir>/script.json (cwd falls back to CLAUDE_CWD). The server reads that file and spawns $SHELL -lc "<command>" in the script's cwd. An out-of-range index (or a missing/invalid script.json) yields { "type": "error", "message": string } and the socket closes.

The output / input / resize / exit frames are identical to /ws. There is no session frame.

Disconnect — the terminal is ephemeral: when the socket closes (cell closed, or page reloaded) the process is killed. There is no background survival and no resume.

Socket.IO: /ws/pubsub (activity pub/sub)

A minimal Socket.IO pub/sub for live session-activity updates. Channel names are Socket.IO rooms.

• Path: /ws/pubsub, transport: websocket.
• Client → server events:
  • subscribe with a channel name (string) → join the room.
  • unsubscribe with a channel name (string) → leave the room.
• Server → client event: data with { channel: string, data: <payload> }.

Channel "sessions" — payloads describe a single session change:

// activity change (working/waiting flipped)
{ "id": "d16f43f3-...", "working": false, "waiting": true, "event": "Stop" }

// a brand-new session was created
{ "id": "…", "working": false, "event": "created" }

// a session's PTY was closed/reaped
{ "id": "…", "working": false, "event": "closed" }

event is the originating hook (UserPromptSubmit | Stop | Notification) or a lifecycle marker (created | closed | null). The client treats any sessions message as a signal to refetch GET /api/sessions (the server is the single source of truth for the list), so payload details are advisory.

---

Session model

Per-session state lives on the server (activity map) and is surfaced as two booleans on every session record:

Flag	Set when	Cleared when	UI
working	UserPromptSubmit hook fires (Claude started a turn)	Stop hook fires (turn finished)	Blue dot next to the title
waiting	A background session fires Notification (waiting for input — permission / question / idle) or Stop (finished, output unseen, ready for another message)	The session is brought to the foreground (a WebSocket attaches to it)	Bold title

"Foreground" = a session that currently has an attached terminal WebSocket (the one you're viewing). waiting is only ever set for background sessions, because a foreground session is already on screen.

---

Session lifecycle

        new ws /ws                         ws /ws?session=<id>
            │                                      │
            ▼                                      ▼
   generate UUID, spawn               live bg PTY?  ──yes──►  reattach + replay buffer
   claude --session-id <uuid>              │ no
   register "New session",                 ▼
   publish "created"               spawn claude --resume <id>
            │                                      │
            └───────────────┬──────────────────────┘
                            ▼
                   attached (foreground)  ── setWaiting(false) ──► not bold
                            │
              ws close (switch away / disconnect)
                            │
            ┌───────── working? ──────────┐
           yes                            no
            │                             │
   keep PTY alive (background)        kill PTY (reap), publish "closed"
            │
   Stop hook in background:
   waiting=true (bold), working=false, reap PTY
   (flag persists via on-disk record → stays listed & bold until viewed)

Key rules:

• Switching away never interrupts Claude mid-turn — a working session's PTY survives in the background.
• A background session that goes idle (Stop) is reaped (killed). If it finished with unseen output, its waiting flag persists via the on-disk session record, so it stays listed and bold until you open it.
• Reattach over respawn: selecting a session that still has a live background PTY reattaches to it (replaying a ≤ 64 KB output tail) instead of spawning a duplicate claude.
• One live viewer per session: a session is bound to a single socket. Opening it in a second place (another tab, or another grid cell pointed at the same dir) reattaches there and supersedes the first, which detaches. To avoid doing this by accident, a grid launcher's resume list flags rows already open in another terminal (● open) and asks for confirmation before taking one over.
• Brand-new sessions appear in the sidebar immediately (before their .jsonl exists) via the in-memory knownSessions registry + a created push; an unused one disappears when its PTY is reaped.
• Background workers get their own filter. A session nobody started by hand — a collection's scheduled refresh, or a plugin's spawnBackgroundChat hidden: true — is listed under the Background chip instead of among the chats, so a refresh schedule doesn't fill the history. It stays openable (a MulmoTerminal session is a live terminal, so a row you can't reach is a process you can't stop), and it is put on the same count+age retention as the scheduler's own sessions. The chip appears only when there is one to show. A manual collection Refresh is a normal visible session — unchanged. The marking is persisted (~/.mulmoterminal/background-sessions.json), so a worker stays out of the chat list after it finishes and after a restart.

---

Claude hook injection

Activity is detected via Claude Code hooks injected per spawn, without touching the user's ~/.claude/settings.json or project settings. The server passes claude --settings '<json>' where the JSON registers a command hook for UserPromptSubmit, Stop, and Notification, each of which pipes the hook payload to the server:

{
  "hooks": {
    "UserPromptSubmit": [{ "hooks": [{ "type": "command", "command": "curl -s -X POST http://localhost:$PORT/api/hook -H 'content-type: application/json' -d @-" }] }],
    "Stop":             [{ "hooks": [{ "type": "command", "command": "curl … -d @-" }] }],
    "Notification":     [{ "hooks": [{ "type": "command", "command": "curl … -d @-" }] }]
  }
}

Because the server spawns each new session with --session-id <uuid>, it always knows the live session's id — even before the session's .jsonl file exists.

---

Closing summary

Every Claude session is spawned with claude --append-system-prompt '<text>', asking the agent to end a reply with a short summary when it hands control back — the work is finished, or it is stopping to ask a question. Coming back to a grid cell after a while, the standing request and what came of it are otherwise only recoverable by scrolling the whole session.

The summary states three things: the request for the conversation as a whole (not the last message — several turns of refinement do not replace what was asked first), what was achieved, and what was not and why. It is written in the language of the conversation, and placed last with nothing after it.

It is deliberately not written on every turn: mid-work replies and short factual answers carry no standing request, and a summary that always appears stops being read. The wording lives in server/agents/session-summary-prompt.ts.

On by default, and switchable off with appendSystemPrompt: false — in ~/.mulmoterminal/config.json, or in a directory's .mulmoterminal.json, which outranks the global value. Read per spawn, so no restart is needed; a session already running keeps what it was launched with. Nothing in the app parses what the summary says, so turning it off costs no feature — the roster and push notifications simply show the raw tail of the reply.

Which sections --append-system-prompt ends up carrying is decided in server/agents/appended-prompt.ts: this one and the prWorkdirFooter clone line are separate settings on the same flag, and with both off the flag is not passed at all.

Passed inline rather than as --append-system-prompt-file for the same reason --settings is: the sandbox spawn runs in a container that cannot read a host path.

Codex sessions are unaffected — the CLI has no equivalent flag.

---

Session discovery & titles

Claude stores each project's sessions as JSONL files under ~/.claude/projects/<encoded-cwd>/<session-id>.jsonl, where the absolute cwd has