Files
AgentHub/server/lib/tmux.js
T
nntrivi2001 ca821008e2 perf: v2.5.10 — ready for 10+ concurrent users, leaner locally
Measured on a copy of a real 1.2 GB / 253K-event database, 10 simulated
browsers + 10 hooks/s for 90 s: event-loop delay p99 1239 ms → 32 ms (max
2.0 s → 94 ms), 266 failed requests → 0, 74 → 111 requests/s served, server
RSS 633 → 346 MB. Entry bundle 975 KB raw → 394 KB (106 KB brotli).

Server
- /api/analytics: the events-table scans run in the read worker
  (scoped-stats eventAggregates, read-pool job analyticsEvents); runRead takes
  maxAgeMs — analytics 10 s, /api/workflows 30 s — keyed by the args
- /api/workflows/skill-usage: +event_type and a tool_name range instead of
  LIKE, so it uses idx_events_tool (784 ms → a few ms)
- idx_events_session_type_tool (session_id, event_type, tool_name,
  created_at) replaces idx_events_session_type: covering reads for session
  stats, the hook dedupes and the import resume point (1.2 s → 20 ms); 64 MB
  cache_size and 256 MB mmap_size on both connections
- WAL: the server sets wal_autocheckpoint = 0 and checkpoints PASSIVE every
  DASHBOARD_WAL_CHECKPOINT_MS (5 s) from the read worker's own writable
  connection; a failed checkpoint restores the automatic one (hook COMMIT p95
  109 ms → 15 ms at 100 hooks/s)
- tmux: one list-sessions answers a whole GET /api/lanes and the heal beat
  (liveSessions), shared for 1 s and cleared by this process's new/kill —
  GET /api/run's sync spawns had blocked 60 s of a 90 s run; the heal beat
  skips tmux when nothing runs; attach options chained into 2 spawns
- hooks: TurnDuration/APIError deduped against one read per hook instead of a
  query per transcript entry (23 ms/hook on a 763-turn session); one lane
  scan per hook; hourly upserts skip rows that did not grow; frames are sent
  after the transaction commits (and after the pane binding); new_event
  carries its row id; a split pane's (slot 1–9) hooks no longer rebind the
  lane's session_id
- watchdog liveness probe runs ps asynchronously; transcript newline scan via
  indexOf; startup subagent backfill fingerprints the files it already checked
- pty attach: backpressure (pause past 1 MB unsent, resume under 256 KB)
- static: hashed assets served from their pre-compressed .br/.gz sibling

Client
- hidden tabs neither refetch on WS events nor poll (useLiveRefresh, panes,
  lane cards, git header, changes panel, settings); no per-tab hourly
  /updates/check (the server already checks)
- refetch storms throttled (workspace tree/session list 3 s, /api/run only on
  a run-key change, notification pane lookup only when shown); WS reconnect
  jitter; terminal retry backoff 1.5 → 30 s, stops on 1008
- heavier throttles for aggregates (15 s) and detail views; Monitor keepIfSame;
  GET in-flight dedupe in request()
- lazy Dashboard/History, xterm (with the first terminal), lane modals, Sagi
  (only while enabled); only the active locale loads; Workspace chunk
  prefetched on a /workspace deep link
- Sagi: no work while disabled, pupil via CSS variables, memo avatar
- build: .br/.gz siblings, sourcemap hidden; service worker no longer caches
  /assets/ (cache v3 drops the old one)

Tests: scripts/test-server.js drops the GIT_DIR-family variables a hook
exports, so a suite run from a linked worktree's pre-commit can no longer
commit into the real repository or set core.bare.

Docs: ARCHITECTURE, DATABASE, API, DEPLOYMENT, client README,
docs/plans/perf/README.md (audit, checklist, results).
2026-10-07 13:24:15 +07:00

368 lines
13 KiB
JavaScript

/**
* @file tmux.js
* @description Thin wrapper around the `tmux` CLI for the terminal-run
* feature. Every dashboard-managed session is named `agenthub-lane-<id>` (see
* `pty-run.js`) so a real terminal can attach to the exact same session with
* `tmux attach -t agenthub-lane-<id>` (or `agenthub lanes shell`). Never builds a
* shell string — every call is `execFileSync("tmux", [...argv])` with an
* explicit argument array (matches this repo's rule for git in worktree.js).
* No call composes a shell command line at all, `sendCommand` included: it
* takes over an idle pane with `respawn-pane -- <argv>` rather than typing a
* quoted line into that pane's shell, which keeps an API key passed in `env`
* out of the pane's rendered text, its scrollback and the user's shell history.
* `sendText` is the one primitive that does send keystrokes — a pane running an
* interactive chat program that reads stdin as a message, not a shell.
* @author Nguyễn Ngọc Trí Vĩ <vinnt@smartgift.vn>
*/
const { execFileSync } = require("node:child_process");
// Test seam: swap the exec implementation so unit tests never invoke a real
// tmux binary. Mirrors run-spawner.js's __injectChildForTest/__reset style.
// stderr is ignored, not inherited: every caller already wraps these in
// try/catch and treats failure as "no session" — piping tmux's own
// "error connecting to ..." noise to the process's stderr just spams the
// journal on every poll while no tmux server is running yet.
const EXEC_OPTS = { encoding: "utf8", stdio: ["pipe", "pipe", "ignore"] };
let execImpl = (args) => execFileSync("tmux", args, EXEC_OPTS);
// Whether this tmux takes `=name` targets: true for tmux itself, false for
// psmux (the WinGet `tmux` on Windows), null until a live session has shown
// which. A fake exec in a test is real tmux unless it says otherwise.
let exactTargets = null;
function __setExecImpl(fn, { exact = true } = {}) {
execImpl = fn;
exactTargets = exact;
listed = null;
}
function __reset() {
execImpl = (args) => execFileSync("tmux", args, EXEC_OPTS);
exactTargets = null;
listed = null;
}
// The `list-sessions` answer, shared for LIST_TTL_MS. Every open dashboard
// polls GET /api/run, and each synchronous spawn forks this whole process —
// measured 2,247 calls, 60 s of a 90 s run blocked, at 10 browsers. A session
// this process creates or kills clears it, so only a pane that exits on its
// own can read as live for up to a second. "No server" (exit 1) is an answer
// too and is shared; any other failure is not.
const LIST_TTL_MS = 1000;
let listed = null; // { at, out } | { at, err }
function listOutput() {
if (listed && Date.now() - listed.at < LIST_TTL_MS) {
if (listed.err) throw listed.err;
return listed.out;
}
try {
const out = String(execImpl(["list-sessions", "-F", "#{session_name}"]));
listed = { at: Date.now(), out };
return out;
} catch (err) {
listed = err.status === 1 || err.code === "ENOENT" ? { at: Date.now(), err } : null;
throw err;
}
}
/**
* Probe once, on a session that exists: `=name` failing where the bare name
* answers means this tmux has no exact-match syntax (psmux answers
* "no server running on session '=name'"), so every target there is the bare
* name. Until a session exists to ask, tmux's own syntax.
*/
function exact() {
if (exactTargets !== null) return exactTargets;
let first;
try {
first = String(execImpl(["list-sessions", "-F", "#{session_name}"]))
.split("\n")
.map((x) => x.trim())
.find(Boolean);
} catch {
return true;
}
if (!first) return true;
const answers = (t) => {
try {
execImpl(["has-session", "-t", t]);
return true;
} catch {
return false;
}
};
if (answers(`=${first}`)) exactTargets = true;
else if (answers(first)) exactTargets = false;
return exactTargets ?? true;
}
// tmux resolves a bare `-t agenthub-lane-1` by PREFIX when no exact match
// exists, so it would hit `agenthub-lane-12` once lane 1's session is gone —
// and killSession would kill the wrong lane. `=` forces an exact match; pane
// and option targets need the trailing `:` (tmux 3.4 rejects `=name` there).
// psmux has neither form and names a session by its bare name.
const session = (name) => (exact() ? `=${name}` : name);
const pane = (name) => (exact() ? `=${name}:` : name);
function hasSession(name) {
try {
execImpl(["has-session", "-t", session(name)]);
return true;
} catch {
return false;
}
}
/** The running tmux server's pid, or null when none answers. */
function serverPid() {
try {
const pid = String(execImpl(["display-message", "-p", "#{pid}"])).trim();
return /^\d+$/.test(pid) ? pid : null;
} catch {
return null;
}
}
/** `-e NAME=value` pairs for tmux, in a stable order. */
function envFlags(env) {
const flags = [];
for (const [key, value] of Object.entries(env || {})) flags.push("-e", `${key}=${value}`);
return flags;
}
/**
* Create a detached tmux session running `argv` as the pane's command. Throws
* if tmux itself fails to start (caller decides how to surface that).
*
* `env` becomes tmux `-e NAME=value` pairs, which tmux puts in the pane's own
* environment (tmux >= 3.2). The dashboard uses it for AGENTHUB_LANE_ID so a `agenthub`
* call from inside the pane knows exactly which lane it belongs to instead of
* guessing from cwd — see resolveLaneArg in bin/agenthub.js.
*/
function newSession({ name, cwd, argv, env }) {
// No argv = tmux's own default shell (a Windows Terminal pane).
const cmd = argv.length ? ["--", ...argv] : [];
listed = null;
execImpl(["new-session", "-d", "-s", name, "-c", cwd, ...envFlags(env), ...cmd]);
try {
enableMouse(name);
} catch (err) {
// The pane's command exited at once (a CLI that refuses the terminal tmux
// gives it, say) and tmux closed the session with it: say that, not the
// "Command failed: tmux set-option …" that hid the real problem.
if (hasSession(name)) throw err;
throw Object.assign(
new Error(
`the session's command exited as soon as it started: ${argv[0] || "the default shell"}`
),
{ code: "EPANEEXITED" }
);
}
}
// Without this the pane has no scrollbar/drag scroll at all — wheel events
// just pass through to the running program instead of entering tmux's own
// copy-mode scrollback. xterm.js forwards tmux's mouse-tracking escapes
// automatically once mouse mode is on, so no client-side change is needed.
// Called on every attach (not just at creation) so a session started before
// this option existed — or before a server restart picked up the change —
// still gets it; `set-option` is idempotent, so re-running it is harmless.
// The status bar goes off here too: the dashboard's tab bar already names the
// pane, and tmux's green line only eats a row of the terminal.
function enableMouse(name) {
// tmux runs `;`-separated commands from ONE argv (each tmux call is a
// synchronous spawn on the event loop), stopping at the first that fails —
// so the options every tmux knows go first, in their own call. psmux (no
// `=name` targets) is not known to chain, so it gets one call per option.
const run = (...cmds) => {
if (exact()) return execImpl(cmds.flatMap((c, i) => (i ? [";", ...c] : c)));
for (const c of cmds) execImpl(c);
};
run(
["set-option", "-t", pane(name), "mouse", "on"],
["set-option", "-t", pane(name), "status", "off"]
);
// A draggable scrollbar beside the pane (mouse mode makes it drag). Only a
// tmux SERVER of 3.7+ knows the option; an older one refuses it, which is
// fine — the pane just has none (`/agenthub-doctor` says how to upgrade).
try {
run(
["set-option", "-t", pane(name), "pane-scrollbars", "on"],
// tmux's default (`fg=white,bg=black,width=1`) is a one-column track in
// ANSI black — the terminal theme's border colour, almost the background —
// so it reads as no scrollbar at all. Bright slider. ONE column: tmux 3.7c
// reserves `width` columns but paints only the first, so a second one is
// never cleared and shows stale cells — a ghost second scrollbar.
[
"set-option",
"-t",
pane(name),
"pane-scrollbars-style",
"fg=brightwhite,bg=black,width=1,pad=0",
]
);
} catch {
/* tmux < 3.7 */
}
}
/**
* The command currently running in the session's active pane (`bash`, `zsh`,
* `claude`, …). Null when tmux can't answer — callers treat that as "unknown,
* don't touch the pane".
*/
function paneCommand(name) {
try {
return (
execImpl(["display-message", "-p", "-t", pane(name), "#{pane_current_command}"]).trim() ||
null
);
} catch {
return null;
}
}
/**
* Run `argv` in an existing session's pane. Only ever called when the pane sits
* at a shell prompt (see `paneCommand`), so what `-k` kills is an idle shell.
*
* `respawn-pane` and not `send-keys`, because `env` may carry a secret — an AI
* Router's API key, which Codex reads from the variable its `env_key` names.
* Typing it would put a `KEY='sk-live-…' codex …` line into the pane's rendered
* text, which the dashboard streams to every attached browser, into the tmux
* scrollback, and into the user's shell history. `-e` hands the value to the
* tmux server through argv instead, exactly as `newSession` already does, and
* the pane never renders it. Running argv directly also means no shell to quote
* for, so the whole class of quoting bugs goes with it.
*/
function sendCommand(name, argv, env, cwd) {
execImpl([
"respawn-pane",
"-k",
"-t",
pane(name),
...(cwd ? ["-c", cwd] : []),
...envFlags(env),
"--",
...argv,
]);
}
/**
* Type a literal line of plain text into an existing pane and press Enter —
* for a program that reads stdin as a message rather than a shell command
* (an interactive `claude`/`codex` chat pane), where `sendCommand`'s POSIX
* quoting and `KEY=value ` env prefix would show up as literal punctuation in
* the message instead of being interpreted. `text` must be single-line: `-l`
* sends it as one literal keystroke burst, so an embedded newline would submit
* only the first line and leave the rest sitting unsent in the input box.
*/
function sendText(name, text) {
execImpl(["send-keys", "-t", pane(name), "-l", text]);
execImpl(["send-keys", "-t", pane(name), "Enter"]);
}
/**
* Paste `text` into the pane as one bracketed paste, then press Enter. Unlike
* `sendText` it may span lines: the text goes into a named buffer through an
* argv element (`set-buffer -- <text>`), never a shell, and `-d` deletes the
* buffer after the paste so a prompt does not linger in tmux's buffer list.
*/
function pasteText(name, text) {
const buf = `agenthub-paste-${name}`;
execImpl(["set-buffer", "-b", buf, "--", text]);
execImpl(["paste-buffer", "-p", "-d", "-b", buf, "-t", pane(name)]);
execImpl(["send-keys", "-t", pane(name), "Enter"]);
}
/** Idempotent — a session that's already gone is not an error. */
function killSession(name) {
listed = null;
try {
execImpl(["kill-session", "-t", session(name)]);
} catch {
/* already gone */
}
}
/** Session names starting with `prefix`. Empty array if tmux has no server running at all. */
function listSessions(prefix) {
let out;
try {
out = listOutput();
} catch {
return [];
}
return out
.split("\n")
.map((s) => s.trim())
.filter((s) => s && s.startsWith(prefix));
}
/**
* Every live session name from ONE `list-sessions` — what a poll asks instead
* of a synchronous has-session per name, each of which blocks the event loop
* (and so every terminal keystroke) for a whole tmux spawn. Empty when no
* server runs (tmux exits 1) or tmux is not installed; null when the call
* itself failed (EAGAIN/ENOMEM under memory pressure), which proves nothing
* about the panes — a caller must not read that as "every pane is gone".
*/
function liveSessions() {
try {
return new Set(
listOutput()
.split("\n")
.map((s) => s.trim())
.filter(Boolean)
);
} catch (err) {
return err.status === 1 || err.code === "ENOENT" ? new Set() : null;
}
}
function isTmuxAvailable() {
return clientVersion() !== null;
}
/** The tmux binary's version (`3.7c`, `next-3.8`), null when it is absent. */
function clientVersion() {
try {
return execImpl(["-V"])
.trim()
.replace(/^tmux\s+/, "");
} catch {
return null;
}
}
/** The RUNNING server's version, null when none runs. A server started before
* an upgrade keeps its old version until every tmux session has closed. */
function serverVersion() {
try {
return execImpl(["display-message", "-p", "#{version}"]).trim() || null;
} catch {
return null;
}
}
module.exports = {
session,
pane,
hasSession,
newSession,
enableMouse,
paneCommand,
sendCommand,
sendText,
pasteText,
killSession,
listSessions,
liveSessions,
isTmuxAvailable,
clientVersion,
serverVersion,
serverPid,
__setExecImpl,
__reset,
};