One entry per command — stable id, plain-language label, key chord, run()
— replacing the ad-hoc keydown listeners in main.ts, catalog.ts, debug.ts
and attach.ts and the parallel SHORTCUTS constant in help.ts. The `?`
overlay now renders the registry, so a binding and its documentation
cannot disagree.
Keys are canonicalised the way the TUI's `InputKey::canonical` spells
them, and chords resolve like its `ChordState`: hold while the sequence
is a strict prefix, otherwise dispatch. That gives the web a leader,
`ctrl-a` by default like `[tui].leader`, held as a partial chord and
`preventDefault`ed so it stops being select-all — the escape hatch in
docs/COMMAND-BAR.md needs somewhere to land (#149). The leader stays out
of real form fields, where select-all is still what the operator wants.
Same keys, same effects: plain single-key bindings remain deaf inside
form fields and the terminal's hidden textarea, chorded ones do not.
Refs #148
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
The pane waited 150 ms on hover and 60 ms on a keyboard step before
following the sidebar highlight, so every preview paid for the fast walk
it might have been. Replace both with a leading-edge 120 ms gap: the
first move shows at once, a burst is coalesced to its start and its end.
The row's own hover no longer eases either — `background` was on a 120 ms
transition, which is what read as lag on the row itself.
The sidebar also rebuilt its whole tree on every `SessionUpdated`, a
couple of times a minute per live session, dropping `:hover` from the row
under the pointer. Refreshes that change nothing structural now patch the
activity rail and meta line into the standing rows instead.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Specifies the ctrl-k command bar for both clients: opening and the
leader escape hatch for a literal Ctrl-K, sigils, empty-query recency,
the key model including the session action page, and touch entry.
The match-and-rank algorithm is pinned exactly — normalization, word
boundaries, the per-character score, embedding choice, tie-break — and
closed with twelve fixtures the TypeScript (#146) and Rust (#147)
matchers are both tested against, so their orders cannot drift.
Insert mode bound `esc` to LeaveInsert, so full-screen apps in a session
(vim, Claude Code) never saw Escape — the only way out of Insert is now
the leader, which rebinding still follows.
A key bound to `Noop` in Insert mode is also forwarded to the pty now.
`Noop` documents itself as the way to suppress a default binding, and in
Insert mode an unbound key belongs to the session, so this is the escape
hatch for taking any other default back.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Two mounts were taking focus out of the sidebar mid-walk: the render that
put the multi-view on screen before the preview applied, and the render
that restores the previous pane when a preview ends. Both now mount
without focusing, so j/k keeps walking. Also marks the preview with a
badge on the canvas and drops the close button from its tab.
The sidebar highlight now shows its session in the pane after a 120ms
debounce, attaching if nothing already is, without the tab taking it on
and without reporting focus to the daemon. Preview-only attaches stay
warm behind the current one, bounded at three.
docs/PREVIEW.md.
Moving the sidebar highlight (keyboard or hover) now shows that session in
the content pane without opening it: its own tab comes forward, or a
throwaway tab carries it. The layout, the saved layout and the daemon's
focus report are all left untouched, and the keyboard stays in the list.
Contract in docs/PREVIEW.md.
Spawn was `s`/`S` in the TUI and `n`/`N` in the web client; the TUI now
uses `n`/`N` too. The web sidebar gains `c`, `r` and `x` — collapse,
rename, kill — which the TUI already bound.
docs/SIDEBAR.md now carries the shared table, so a new binding has one
place to land on both clients.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
The TUI repainted on every 33ms tick whether or not anything had
changed — measured at ~725 bytes/s of escape sequences with nothing
attached and nothing happening. Over a remote terminal that is constant
traffic for no visible change.
Output was also unbuffered: `CrosstermBackend::new(std::io::stdout())`
issues one `write!` per cell run, 3551 of them for a 200x50 repaint,
each taking the global stdout lock. A BufWriter collapses a frame into
one flush (measured: 3551 writes -> 2).
Frames are now keyed on the catalog version, the active pane's screen
revision, an input epoch and the terminal size, and skipped when the key
is unchanged. Session cards carry an advancing duration, so a frame
still never stands for longer than a second.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Adding @xterm/addon-webgl changed web/pnpm-lock.yaml, which the flake's
pnpmDeps.hash pins against, so `nix build .#webFrontend` failed offline
on the missing tarball and took the CI flake job with it.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
The dev daemon ran with the toolchain PATH from `nix run .#dev`, whose
trailing `/usr/bin:/bin` is all but empty on NixOS. Sessions spawned
from the dev UI inherit that PATH, so every one came up unable to find
`claude`, `zsh`, or anything else installed in the user's profile.
Append the usual profile directories in the daemon's command, resolved
by the shell at run time. Build tools still come first; directories that
do not exist on a given host cost nothing.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
xterm's DOM renderer builds a node per cell, which is CPU-bound — the
wrong bottleneck on a phone, where an agent dumping output stutters
visibly on a Pixel 7 despite plenty of GPU headroom.
A WebGL context is scarce and revocable, so it follows the mounted view
rather than the connection: acquired in onMount, released in onUnmount.
ConnectionPool keeps MAX_PARKED connections alive behind the visible
ones, which would otherwise exceed the browser's context cap (~16 on
desktop Chrome, fewer on Android) and bring a parked terminal back
blank. Context loss falls back to the DOM renderer and retries against a
shared two-loss budget; a device that keeps dropping contexts stops
being asked. A device pixel ratio change rebuilds the glyph atlas, which
is otherwise left blurry.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
The dialog capped at 92vh and scrolled its whole body, so the Spawn button
sat below the fold on a phone. Below 600px it now fills the viewport, the
pickers drop their own scroll caps, and the actions row sticks to the bottom
edge.
The daemon's auto-install skipped on presence, so a hook installed
before #131 never gained UserPromptSubmit and no session was ever named
after its prompt.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
The TUI arm of the umbrella binary hardcoded http://127.0.0.1:7070 instead
of going through MaestroConfigFile::resolve_daemon_url, which every CLI
subcommand uses. On a host whose [daemon].bind_addr is not the default port,
`maestro list` worked and `maestro tui` could not connect.
RunConfig.daemon_url is now the override rather than the final URL; the TUI
resolves it against the config file it already loads for its UI settings.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Keystrokes and PTY echo are tiny frames. With Nagle on, each can sit up to
40ms waiting for an ACK before the kernel sends it — invisible over
loopback, plainly felt from another device on the LAN.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
The browser reached the daemon through maestro-web, a second binary that
served the SPA and reverse-proxied /api/* — two TCP connections and a
WebSocket frame re-encode on every keystroke and every byte of PTY output.
On one machine that hop buys nothing.
The daemon now embeds the bundle and serves it off its own listener, behind
a default-on `web` feature. `--no-default-features` builds without pnpm or
node; MAESTRO_WEB_DIST still takes a prebuilt bundle.
The router's fallback answers an unmatched /api/* with 404 rather than the
SPA shell, so a client expecting JSON gets a status it can read instead of
HTML.
Removes maestro-web and the [web] config section. An existing [web] block
is ignored, not rejected. The installer deletes a leftover
maestro-web.service on upgrade.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Skipping the operator's MCP servers, settings and hooks takes a measured
naming run from 12s to 7s, so the timeout is a margin rather than a race.
Refs #131
The name field starts empty behind a hint, ^R reroll is gone, and the
fetched handle seeds the worktree branch instead. Fast-forward spawns with
no name at all. A card whose name the daemon generated renders muted.
Refs #131
A session's rail glows AwaitingInput even after the operator has already
looked at it. Clients now report pane focus (POST /:id/focus); the daemon
clears the glow to Idle after 5s of continuous focus and keeps it cleared
until the session does new work, so a fresh Active cycle is required to
glow again.
The name field starts empty and the reroll is gone; the worktree branch keeps
a default by borrowing one handle from /api/names. A card whose name the
daemon generated renders dimmed until a first prompt settles it.
Refs #131
Tab now moves the keyboard from the session list to the pane (was
Escape); Escape moves it back from the pane to the list, intercepted via
xterm's attachCustomKeyEventHandler so it never reaches the shell/app.
Matches the TUI's `tab`/`esc` pair; Ctrl+Q was ruled out since browsers
reserve it and never deliver it to page JS.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Normal mode only exists while the sidebar is focused (Insert mode always
pairs with Focus::Pane), so bare `q` and `tab` there never reach a pty:
`q` quits, `tab` focuses the highlighted pane (same as Enter). `esc` is
added as a second, fixed way out of Insert alongside the leader.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
POST /api/sessions/:id/prompt claims a session's one naming attempt and runs
claude -p --model haiku in the background; the answer is reduced to a short
slug and replaces the generated handle, settling the session at Haiku. A
chosen name, a rename mid-run, a missing binary or a timeout all leave the
handle alone. The prompt goes to the child on stdin and is never logged.
Refs #131
Adds a UserPromptSubmit entry that forwards the hook payload verbatim to
/api/sessions/:id/prompt, gated by a haiku-autoname feature that the daemon,
the CLI and the umbrella binary all forward. With it off no hook points at
the route.
Refs #131
The spawn form no longer prefills a name, so create() fills one in: a
requested name is marked User, an absent one gets a generated handle marked
Generated. Rename always settles to User.
Refs #131
The `flake` job was the pipeline's critical path at ~9 minutes, all of it
one cold `nix build`: with an empty /nix every run, it recompiled the whole
release closure under `lto = "thin"` + `codegen-units = 1`.
- Cache /nix across runs, keyed on the lockfiles the closure derives from,
so crane's deps-only derivation survives and only maestro's own crates
rebuild. Saved from pushes only, so a PR cannot write a partial store.
- Build only `.#webFrontend` on PRs — that is the step that guards
pnpmDeps.hash. The release binaries stay on main.
- Replace the hand-rolled `target/` cache with Swatinem/rust-cache, which
prunes to the current lockfile instead of accumulating stale artifacts
behind a prefix restore-key. ~60s per run went into moving that entry.
- Run tests under cargo-nextest, which parallelises across the 20
integration binaries. Capped at 4 threads in .config/nextest.toml: the
tmux-backed tests starve each other at the num-cpus default and trip
their 5s deadlines.
- actions/setup-node instead of a second apt-get pass via nodesource.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Adds NameSource (User/Generated/Haiku), the metadata keys behind it, the
POST /api/sessions/:id/prompt body, and display helpers both sidebars will
read to decide whether a name is provisional.
Refs #131
Sidebar rows only navigated via location.hash + hashchange. Moving
pane focus by split, drag, or direct pane click never touched the
hash, so a later click back onto a session whose row href matched the
stale hash fired no hashchange and was silently ignored — the pane
had to be clicked directly to regain focus.
Row clicks now also dispatch a session-select event handled the same
way as a hash deep-link, independent of whether the hash text changes.
N picked the globally freshest project by git-mtime, ignoring the session
the operator was actually on. Plain n's spawn form already seeds from the
focused session (spawnSimilarProjectSeed); fast-forward now matches that,
falling back to recency only when nothing is focused.
Rail now spans full row height flush to the left edge with square
corners, instead of a depth-shifted inset bar. Active state stays
solid green since colour alone already signals it; blink/flash stay
for awaiting-input/error.
naps62-yolo