@flame0510/project-aether 1.6.1 → 1.7.0

package/app/components/SessionDrawer.tsx

@@ -180,27 +180,27 @@ export default function SessionDrawer({ sessionId, onClose }: SessionDrawerProps
               {truncate(session.session_id, 60)}
             </div>
             <Grid>
-              <Row label="STATO">
+              <Row label="STATUS">
                 <span style={{ color: statusColor(session.status) }}>●</span>{' '}
                 {session.status ?? 'idle'}
               </Row>
-              <Row label="TIPO">
+              <Row label="TYPE">
                 {session.session_id.includes(':cron:') ? 'cron' : 'session'}
               </Row>
               <Row label="MODEL">{session.model ?? '-'}</Row>
-              <Row label="COSTO">{formatUsdOrDash(session.cost_usd)}</Row>
+              <Row label="COST">{formatUsdOrDash(session.cost_usd)}</Row>
               <Row label="TOKEN">
                 in: {formatTokens(session.tokens_in)} out: {formatTokens(session.tokens_out)}
               </Row>
-              <Row label="INIZIO">{formatTimeFromUnixSeconds(session.started_at)}</Row>
-              <Row label="AGGIORN.">{formatTimeFromUnixSeconds(session.updated_at)}</Row>
-              <Row label="FINE">{formatTimeFromUnixSeconds(session.ended_at)}</Row>
-              <Row label="DURATA">{formatDuration(session.started_at, endedOrUpdated)}</Row>
+              <Row label="STARTED">{formatTimeFromUnixSeconds(session.started_at)}</Row>
+              <Row label="UPDATED">{formatTimeFromUnixSeconds(session.updated_at)}</Row>
+              <Row label="ENDED">{formatTimeFromUnixSeconds(session.ended_at)}</Row>
+              <Row label="DURATION">{formatDuration(session.started_at, endedOrUpdated)}</Row>
               <Row label="TASK">{truncate(session.task_preview ?? '-', 200)}</Row>
             </Grid>
           </Section>
 
-          <Section title="AZIONI RECENTI">
+          <Section title="RECENT ACTIONS">
             {events.length === 0 && <Muted>No events</Muted>}
             {events.map((evt) => (
               <div key={evt.id ?? `${evt.ts ?? 0}-${evt.type ?? 'event'}`} style={{ fontSize: 12 }}>

package/app/components/SystemCockpit.tsx

@@ -144,7 +144,7 @@ export default function SystemCockpit({
                     <span>{item.severity}</span>
                   </div>
                   <div className="ui-muted" style={{ marginTop: 4 }}>{item.details}</div>
-                  {item.actionHref && <Link href={item.actionHref} className="ui-link" style={{ display: 'inline-block', marginTop: 6 }}>Apri →</Link>}
+                  {item.actionHref && <Link href={item.actionHref} className="ui-link" style={{ display: 'inline-block', marginTop: 6 }}>Open →</Link>}
                 </li>
               ))}
             </ul>

package/app/globals.css

@@ -1387,7 +1387,7 @@ html, body { height: 100%; height: 100dvh; background: var(--bg); color: var(--t
     white-space: nowrap;
   }
 
-  /* Padding bottom per non coprire contenuti — handled by .app-shell__content */
+  /* Bottom padding so content is not covered — handled by .app-shell__content */
 
   /* Chat button raised above the bottom navbar */
   .ochat__trigger {

package/app/setup/PageClient.tsx

@@ -149,7 +149,7 @@ export default function SetupPageClient() {
               <div style={{ display: 'grid', gap: 6, textAlign: 'left', padding: '10px 12px', background: 'rgba(52,211,153,0.06)', border: '1px solid rgba(52,211,153,0.15)' }}>
                 {[
                   'Set admin password',
-                  'Agent base image is built by the setup script',
+                  'Agent base image: download it from the banner on the Agents page',
                   'Next: add providers & agents in the wizard',
                 ].map((t) => (
                   <div key={t} style={{ display: 'flex', alignItems: 'center', gap: 8, fontSize: 12, color: 'var(--text)' }}>

package/bin/postinstall.js

@@ -1,6 +1,10 @@
 #!/usr/bin/env node
 /**
  * Post-install hook: build Next.js app if .next doesn't exist.
+ *
+ * Skipped when REV4A_SKIP_POSTINSTALL_BUILD is set: CI and the release workflow run
+ * `npm run build` as an explicit step, and a second build here would run before every
+ * other check and swallow its own failure.
  */
 const { existsSync } = require('fs');
 const { join } = require('path');
@@ -8,7 +12,7 @@ const { join } = require('path');
 const ROOT = join(__dirname, '..');
 const NEXT_DIR = join(ROOT, '.next');
 
-if (!existsSync(NEXT_DIR)) {
+if (!process.env.REV4A_SKIP_POSTINSTALL_BUILD && !existsSync(NEXT_DIR)) {
   try {
     const { execSync } = require('child_process');
     console.log('[rev4a] Building dashboard… (one-time, may take a minute)');

package/daemon.js

@@ -14,7 +14,6 @@ const path = require('path');
 
 const DB_PATH = process.env.REV4A_DB || path.join(os.homedir(), '.config', 'rev4a', 'data', 'events.db');
 const POLL_INTERVAL_MS = 30_000;
-const POLL_INTERVAL_ACTIVE_MS = 15_000; // 15s when working sessions detected
 
 // Cost rates per 1M tokens (separate in/out pricing)
 const MODEL_PRICING = {
@@ -300,19 +299,14 @@ function readHostCpuSample() {
   return { idle, total };
 }
 
+// The filesystem holding Rev4a's own data (the events database's directory), in GB.
+// statfs works the same on Linux and macOS; the previous `df -BG /data` read a mount
+// that existed only on one old server and used GNU-only flags, so it reported 0/0.
 function getDiskStats() {
   try {
-    const result = spawnSync('df', ['-BG', '/data', '--output=used,size'], {
-      encoding: 'utf8', timeout: 5000, killSignal: 'SIGKILL',
-    });
-    if (result.status !== 0 || !result.stdout) return { used: 0, total: 0 };
-    const lines = result.stdout.trim().split('\n');
-    // lines[0] = header, lines[1] = data
-    if (lines.length < 2) return { used: 0, total: 0 };
-    const parts = lines[1].trim().split(/\s+/);
-    const used = parseFloat(parts[0]) || 0;  // already in GB (BG flag strips G)
-    const total = parseFloat(parts[1]) || 0;
-    return { used, total };
+    const st = fs.statfsSync(path.dirname(DB_PATH));
+    const gb = (blocks) => Math.round((blocks * st.bsize) / 1024 ** 3);
+    return { used: gb(st.blocks - st.bfree), total: gb(st.blocks) };
   } catch (e) {
     return { used: 0, total: 0 };
   }
@@ -357,10 +351,10 @@ function collectSystemMetrics() {
             metric: 'cpu',
             values: [recent[1].cpu_percent, recent[0].cpu_percent],
             threshold: 85,
-            message: `CPU alta: ${recent[0].cpu_percent}% per 2 campioni consecutivi`,
+            message: `CPU high: ${recent[0].cpu_percent}% for 2 consecutive samples`,
           }),
         });
-        log(`[ANOMALY] CPU alta: ${recent[0].cpu_percent}%`);
+        log(`[ANOMALY] CPU high: ${recent[0].cpu_percent}%`);
       }
     }
     // RAM >90% for 2 consecutive samples
@@ -377,10 +371,10 @@ function collectSystemMetrics() {
             metric: 'ram',
             values: [ram1pct, ram0pct],
             threshold: 90,
-            message: `RAM alta: ${ram0pct}% per 2 campioni consecutivi`,
+            message: `RAM high: ${ram0pct}% for 2 consecutive samples`,
           }),
         });
-        log(`[ANOMALY] RAM alta: ${ram0pct}%`);
+        log(`[ANOMALY] RAM high: ${ram0pct}%`);
       }
     }
   }
@@ -540,7 +534,7 @@ function pollSessions() {
     const session_id = s.key;
 
     // Skip Telegram channel/group sessions (multi-user).
-    // Keep Telegram direct sessions (agent:ops:telegram:argus:direct:...) as root nodes
+    // Keep Telegram direct sessions (agent:<id>:telegram:<account>:direct:...) as root nodes
     // since they are the parent of all sub-agents spawned via Telegram.
     if (session_id.includes(':telegram:') && !session_id.includes(':direct:')) continue;
     const { label, parent_id: inferredParent } = parseSessionKey(s.key);
@@ -698,27 +692,12 @@ function pollSessions() {
       }),
     });
     log(`[TIMEOUT] ${session_id} missing for ${Math.round(missingFor / 60000)} min`);
-
-    // Notify Michele via openclaw message (only if openclaw is available)
-    try {
-      if (!openclawMissing) {
-        execSync('which openclaw', { stdio: 'ignore', timeout: 3000 });
-      }
-      spawnSync('openclaw', [
-        'message', 'send',
-        '--account', 'ops',
-        '--target', '297086793',
-        '--text', `⚠️ Rev4a: agent timeout\n\`${session_id.slice(-36)}\`\nMissing for ${Math.round(missingFor / 60000)} min without completing.`,
-      ], { encoding: 'utf8', timeout: 10_000, killSignal: 'SIGKILL' });
-    } catch (e) {
-      log(`[TIMEOUT] Telegram notification failed: ${e.message}`);
-    }
   }
 
   // Collect system metrics after each poll
   try { collectSystemMetrics(); } catch (e) { log(`[METRICS ERROR] ${e.message}`); }
 
-  // Force names and parents declared via lineage — overrides any previous label (including "Sub-agente")
+  // Force names and parents declared via lineage — overrides any previous label, including the generic placeholder older data carries
   const updateLabel = db.prepare('UPDATE sessions SET label = ?, parent_id = ? WHERE session_id = ?');
   const applyLineage = db.transaction(() => {
     for (const [child_id, agent_name] of Object.entries(declaredNames)) {
@@ -741,7 +720,7 @@ function pollSessions() {
     log(`Retention cleanup: ${r1.changes} cron sessions, ${r2.changes} cron events deleted`);
   }
 
-  // Prune knownSessions — delete completed sessions da >30 giorni
+  // Prune knownSessions — delete completed sessions older than 30 days
   const cutoff = Date.now() - (30 * 24 * 60 * 60 * 1000);
   for (const [id, snap] of knownSessions) {
     if (snap.status === 'completed' && snap.updatedAt && snap.updatedAt < cutoff) {

package/docs/ARCHITECTURE.md

@@ -1,7 +1,7 @@
 # Rev4a Architecture — Design & Vision
 
 > **Status:** Active — `main` branch
-> **Last updated:** 2026-09-22
+> **Last updated:** 2026-09-26
 > **Goal:** Transform Rev4a from a monitoring dashboard into a central orchestrator for a distributed multi-container agency.
 
 ---
@@ -35,7 +35,7 @@
 
 ```
 ┌────────────────────────────────────────────────────────────────┐
-│  DOCKER HOST (187.77.156.41, 4 CPU, 15 GB RAM, 193 GB disk)   │
+│  DOCKER HOST                                                   │
 │                                                                │
 │  ┌──────────────────┐  ┌──────────────────┐  ┌──────────────┐ │
 │  │ rev4a-control  │  │ agent-argus      │  │ agent-atlas  │ │
@@ -132,6 +132,18 @@ The central container, running the Next.js dashboard + orchestration API.
   a key), `isModelOffered()` (enforced by the proxy and the assistant) and
   `catalogueStatus()`. A failed read of `models.config.json` falls back to the last
   good copy and never deletes overrides.
+- **The model data has three files and three scripts.** `models.config.json` is curated by
+  hand (identity, `enabled`, `deprecated`, and an optional `info` block for what no API
+  publishes: the vendor's size claim, benchmarks, the docs link, notes).
+  `model-pricing.json` and `model-details.json` are **generated** by
+  `npm run refresh:pricing` (`scripts/refresh-model-pricing.mjs`): prices from OpenRouter,
+  and per model the description, architecture and benchmarks from OpenRouter plus the size,
+  weight mix, licence and dates from the Hugging Face card. Two scripts back the curation
+  of `info`: `npm run info:suggest` (`scripts/model-info-suggest.mjs`) prints the candidate
+  facts from a card, and `npm run info:verify` (`scripts/model-info-verify.mjs`) refuses a
+  proposal the card does not support — evidence per field, every benchmark value under the
+  column that names this model. `GET /api/models/details` merges the three files for the
+  details modal (`docs/dev/GATEWAY.md`).
 - **Agent images and OpenClaw versions.** `lib/agent-versions.json` lists the supported
   OpenClaw versions, newest first, with the model `input` list for each; the server reads
   it through `lib/agent-versions.ts`, the `rev4a` CLI with `require`. The provider sync
@@ -560,19 +572,18 @@ The Rev4a daemon (`daemon.js`) is a standalone Node.js process that bridges the
 
 ### Responsibilities
 
-1. Poll `openclaw sessions --json --all-agents` on a configurable interval
+1. Poll `openclaw sessions --json --all-agents` every 30 s
 2. Upsert session rows into `events.db`
-3. Emit `spawn` / `complete` / `error` events into the `events` table
-4. Collect system metrics (CPU, RAM, disk) every poll cycle
-5. Detect anomalies (CPU > 90%, RAM > 90%) and log warnings
+3. Emit `spawn` / `complete` / `fail` / `spawn_timeout` events into the `events` table
+4. Collect system metrics (CPU, RAM, disk) after every completed poll
+5. Detect anomalies (CPU > 85%, RAM > 90%) and record them
 6. Manage DB lifecycle (WAL mode, checkpoint after each cycle)
 
-### Poll Intervals
+### Poll Interval
 
-| Condition | Interval |
-|---|---|
-| No active sessions | 30 s |
-| ≥ 1 session with status `working` | 15 s |
+A fixed 30 s timer, whether or not sessions are working. When the OpenClaw CLI is not on
+the host, the daemon logs it once and stays idle — and then records no system metrics
+either, since they are collected at the end of a poll.
 
 ### Cost Estimation
 
@@ -614,28 +625,33 @@ The daemon uses `INSERT … ON CONFLICT DO UPDATE` with these rules:
 
 ### System Metrics
 
-Every poll cycle the daemon records a `system_metrics` row. Data sources:
+After every completed poll the daemon records a `system_metrics` row. Data sources:
 
-- **CPU**: cgroup v2 usage delta (`/sys/fs/cgroup/cpu.stat`) when available, falls back to `/proc/stat` host ticks
-- **RAM**: `/proc/meminfo`
-- **Disk**: `df -BG /data`
+- **CPU**: cgroup v2 usage delta (`/sys/fs/cgroup/cpu.stat`) when available, falls back to host ticks from `os.cpus()`
+- **RAM**: `os.totalmem()` / `os.freemem()`
+- **Disk**: `fs.statfsSync()` on the directory holding the events database — the
+  filesystem Rev4a's own data lives on; works the same on Linux and macOS
 - **Load**: `os.loadavg()[0]`
 
-Metrics older than 24 h are pruned automatically each cycle.
+Metrics older than 30 days are pruned automatically each cycle.
 
 ### Anomaly Detection
 
-The daemon compares the last two metric samples. If a threshold is exceeded and the cooldown (5 min) has passed, it logs a `WARN` line:
+The daemon compares the last two metric samples. When both exceed a threshold and the
+cooldown (5 min per metric) has passed, it inserts a `system_anomaly` event
+(`{ metric, values, threshold, message }`) and logs `[ANOMALY] CPU high: N%` or
+`[ANOMALY] RAM high: N%`. No feature consumes `system_anomaly` events yet; they only pass through the generic
+event feeds.
 
-| Metric | Threshold |
+| Metric | Threshold (two consecutive samples) |
 |---|---|
-| CPU | > 90% |
+| CPU | > 85% |
 | RAM | > 90% |
 
 ### DB Safety
 
 - WAL mode + `PRAGMA synchronous = NORMAL` for concurrent read safety
-- `PRAGMA wal_checkpoint(PASSIVE)` runs after each poll cycle
+- `PRAGMA wal_checkpoint(PASSIVE)` runs after each poll cycle, `FULL` every 10 cycles
 - On `uncaughtException` the daemon logs but does NOT exit — relies on systemd Restart=always for recovery
 
 ### Migrations
@@ -722,17 +738,19 @@ ALTER TABLE sessions ADD COLUMN ended_at INTEGER;
 
 Rev4a provides a real, interactive terminal for any agent container via
 a WebSocket-connected PTY. The implementation is documented in detail in
-[docs/container-terminal.md](container-terminal.md).
+[docs/CONTAINER-TERMINAL.md](CONTAINER-TERMINAL.md).
 
 ### Two-process architecture
 
 | Process | Port | Role |
 |---|---|---|
-| `rev4a-next` (Next.js) | 3740 | Serves the terminal page, auth, API |
-| `rev4a-terminal-ws` (standalone) | 3741 | WebSocket PTY server via `node-pty` |
+| Next.js (`next start`) | 3740 | Serves the terminal page, auth, API |
+| `terminal-ws-server.js` | 3741 (127.0.0.1) | WebSocket PTY server via `node-pty` |
 
-The terminal server runs as a separate systemd service (rev4a-terminal-ws) to avoid event-loop
-contention with Next.js during high-throughput I/O.
+Both are started by `rev4a serve`; the terminal server is a process of its own to avoid
+event-loop contention with Next.js during high-throughput I/O. It has no authentication
+of its own and is reachable only through a reverse proxy route (see
+[CONTAINER-TERMINAL.md](CONTAINER-TERMINAL.md#routing)).
 
 ### Why custom DOM over xterm.js
 
@@ -742,7 +760,7 @@ CSS conflicts. The current implementation uses a plain `<div>` with native
 browser scrolling and a hidden `<textarea>` for input — stable under any
 output volume.
 
-See [container-terminal.md](container-terminal.md#terminal-client-browser)
+See [CONTAINER-TERMINAL.md](CONTAINER-TERMINAL.md#terminal-client-browser)
 for the full rationale.
 
 ## 12. Technology Stack

package/docs/CONTAINER-TERMINAL.md

@@ -15,7 +15,7 @@ via `node-pty` and streaming I/O over a dedicated WebSocket server.
 Browser (xterm-compatible DOM terminal)
   │  WebSocket wss://rev4a/containers/terminal/{id}
   ▼
-terminal-ws-server.js (port 3741, proxied by Traefik)
+terminal-ws-server.js (127.0.0.1:3741, reached through a reverse proxy — see Routing)
   │  node-pty
   ▼
 docker exec -it {containerId} env TERM=xterm-256color bash
@@ -30,20 +30,30 @@ Container Shell (bash, interactive, full PTY)
 
 | Process          | Port  | Role |
 |------------------|-------|------|
-| `rev4a-next`   | 3740  | Next.js app (pages, API routes, auth) |
-| `rev4a-terminal-ws` | 3741 | Standalone WebSocket server for terminal sessions |
+| Next.js (`next start`) | 3740  | Next.js app (pages, API routes, auth) |
+| `terminal-ws-server.js` | 3741 | Standalone WebSocket server for terminal sessions, bound to 127.0.0.1 |
 
-The terminal server is a separate systemd service (rev4a-terminal-ws). It was isolated from
+Both are child processes of `rev4a serve` (with `daemon.js`), under the single
+`rev4a.service` unit in production. The terminal server is its own process, isolated from
 the Next.js app to avoid event-loop contention during high-throughput I/O (fast
 `seq`, `cat` on large files, interactive shell sessions).
 
-### Traefik routing
+### Routing
 
 ```
 /containers/terminal/{id}          → Next.js (page serving TerminalClient)
-/api/terminal-ws?id={containerId}  → WebSocket proxy → ws://127.0.0.1:3741
+/api/terminal-ws?id={containerId}  → reverse proxy → ws://127.0.0.1:3741
 ```
 
+The browser opens the socket on the dashboard's own host (`/api/terminal-ws`). Next.js
+has no route for that path, so the terminal works only when a reverse proxy in front of
+Rev4a forwards it to port 3741; without one the socket gets a 404.
+
+**No authentication of its own.** `terminal-ws-server.js` checks neither the session
+cookie nor the token: whoever reaches it gets a shell in the container named by `id`.
+It listens on 127.0.0.1 only, so the exposure is exactly what the proxy route above
+opens. An open issue.
+
 ### Data flow
 
 1. User opens `/containers/terminal/openclaw-atlas`
@@ -257,8 +267,7 @@ app/containers/terminal/
 ├── [id]/
 │   ├── page.tsx               # Next.js page (auth-protected)
 │   └── TerminalClient.tsx     # Client-side terminal component
-terminal-ws-server.js           # WebSocket PTY server (rev4a-terminal-ws.service)
-rev4a.service                  # Systemd target for all three services
+terminal-ws-server.js           # WebSocket PTY server, started by `rev4a serve`
 ```
 
 ## Known Limitations

package/docs/DESIGN-SYSTEM.md

@@ -2,7 +2,7 @@
 
 Source of truth: `app/design-system.ts` + CSS tokens in `app/globals.css`.
 
-> **Last updated:** 2026-07-03
+> **Last updated:** 2026-09-26
 
 ---
 
@@ -44,19 +44,22 @@ Rev4a uses a **square/sharp** visual language:
 Defined in `app/globals.css` as CSS custom properties:
 
 ```css
---bg:           #0a0a0a      /* Page background */
---bg2:          #111         /* Slightly lighter surface */
---border:       #222         /* Default border */
---text:         #e0e0e0      /* Primary text */
---text-dim:     #888         /* Muted/secondary text */
---violet:       #a78bfa      /* Brand accent */
---violet-bg:    #1a1030     /* Violet-tinted hover/active background */
---green:        #22c55e      /* Success */
---red:          #ef4444      /* Error */
---yellow:       #eab308      /* Warning */
---blue:         #60a5fa      /* Info / links */
+--bg:         var(--color-bg-950)       /* #0A0A0B  Page background */
+--bg2:        var(--color-bg-900)       /* #111114  Slightly lighter surface */
+--border:     var(--color-border-700)   /* #222228  Default border */
+--text:       var(--color-text-100)     /* #E8E8E8  Primary text */
+--text-dim:   var(--color-text-400)     /* #888     Muted/secondary text */
+--violet:     var(--color-accent-500)   /* #925BFC  Brand accent */
+--violet-bg:  rgba(146, 91, 252, 0.08)  /* Violet-tinted hover/active background */
+--green:      var(--color-success-500)  /* #22c55e  Success */
+--red:        var(--color-danger-500)   /* #ef4444  Error */
+--yellow:     var(--color-warning-500)  /* #f59e0b  Warning */
+--blue:       var(--color-blue-400)     /* #60a5fa  Info / links */
 ```
 
+The semantic tokens above point at the palette tokens (`--color-*`) defined at the top of
+`app/globals.css`; the hex values are the palette's, quoted for reference.
+
 ### Typography
 
 - Monospace by default: `var(--font-mono-stack)`
@@ -66,6 +69,12 @@ Defined in `app/globals.css` as CSS custom properties:
 
 ## Component design rules
 
+A shared component grows by **variant, never by copy**. `Metric` is the example: `size="sm"`
+is the compact form for a grid inside a dialog (smaller value, tighter spacing), and it lives
+on the primitive so the modal and the dashboard cannot drift apart. When a new page needs a
+pattern that does not exist yet, add it here and to `docs/FRONTEND-ARCHITECTURE.md`'s catalog,
+then use it — a one-off copy of markup is the thing this section exists to prevent.
+
 1. **Border radius: 0 everywhere.** `border-radius: 0` or sharp corners. Exceptions: `Pill`, avatar circles, status dots.
 2. **No shadows.** Flat design. Use `border: 1px solid var(--border)` to define surfaces.
 3. **Violet is the only accent.** `var(--violet)` for active states, primary actions, headings. `var(--violet-bg)` for hover/active backgrounds.

package/docs/FRONTEND-ARCHITECTURE.md

@@ -1,6 +1,6 @@
 # Rev4a Frontend Architecture
 
-> **Last updated:** 2026-09-22
+> **Last updated:** 2026-09-26
 
 ## Layering
 
@@ -45,7 +45,7 @@ All shared UI primitives live in `app/components/ui/` and are exported from `app
 | `UpdateSection` | `app/agents/UpdateSection.tsx` | OPENCLAW VERSION section of the agent detail panel: the version the agent runs, **Update to <version>** when a newer supported version is downloaded (confirm modal, disabled while the agent is stopped), the running update's steps with backup progress, the outcome, and **Roll back to <version>** after an update. Polls `/api/agents/[id]/update` every 2 s while an update or rollback runs; one action at a time (keyed busy state). |
 | `BackupSection` | inline in `app/agents/PageClient.tsx` | BACKUP section of the agent detail panel, on the cold backup and the restore. **Backup Now** starts `POST /api/agents/[id]/cold-backup`; while the job runs a banner shows the file and its live percent with **Cancel** (`DELETE /cold-backup`). **Restore** (after a confirm) starts `POST /restore` and a banner shows `Restoring <file>…` (no percent: the extract is a single `tar xzf`, and there is no Cancel). Both sections poll their `GET` every 2 s while running, and on mount pick up a job that is already running — a backup lives in a Docker helper, a restore in `agent_restores`, so reloading the page or navigating away never loses them nor allows a second one (the server answers 409 anyway). Delete per row; all actions disabled while one runs; keyed busy state `{ kind, file }` so only the row in action shows the spinner. On the agent list, an activity Badge (fed by `/api/agents/activity-summary`, polled at 2 s only while something runs, otherwise riding the 15 s list poll) reads `BACKUP nn%`, `RESTORING`, `RECREATING`, `EDITING` or `UPDATING`. |
 | `RecreateSection` | inline in `app/agents/PageClient.tsx` | RECREATE section of the agent detail panel. **Recreate Container** starts `POST /api/agents/[id]/recreate` (202) after a confirm; a banner then shows the phase — *Backing up … nn%* while the cold backup runs, *Recreating container…* while the container is rebuilt and the gateway starts. The section polls `GET /recreate` every 2 s, and on mount picks up a recreate that is already running, so a reload or navigation never loses it; it refetches the agent once the job reports `done`. |
-| `ImageDownloadBanner` | `app/agents/ImageDownloadBanner.tsx` | Agent image banner on the Agents page. Polls `/api/agents/image-status` every 2 s; offers **Download Image** when no supported version is downloaded, **Download <version>** when the registry publishes a newer one, and shows the download in progress and its completion. Downloading changes no agent. |
+| `ImageDownloadBanner` | `app/agents/ImageDownloadBanner.tsx` | Agent image banner on the Agents page. Polls `/api/agents/image-status` every 2 s; offers **Download Image** when no supported version is downloaded, **Download <version>** when the registry publishes a newer one, and while a download runs shows the percent of layers finished and the latest line of docker's output (the bar is indeterminate until a percent can be computed). **Cancel** (`DELETE /api/agents/download-image`, own loading state) appears once `image-status` reports the download running; before that the button reads *Starting…* and is disabled. Every outcome — ready and cancelled for 3 s, a failure until the next action — comes from the server's `lastResult`, so one that ended while the page was closed or reloading is still reported if it is less than 30 s old (aged with `serverTime`). Each outcome is announced once per page load, although the Agents page mounts the banner in three places (mobile list, mobile detail, desktop). Downloading changes no agent. |
 | `VersionBanner` | `app/components/VersionBanner.tsx` | "Update available" banner for Rev4a itself, when `GET /api/update-check?check=1` reports a newer published version (dismissable per version, remembered in `localStorage`). **Update now** starts `POST /api/update-check`; because that update restarts the server, the banner cannot be told the outcome by the response: it records what it asked for in `sessionStorage`, polls `/api/update-check` until the installed version moves (two minutes at most) and reloads, then on the next mount either confirms "Updated to vX" or reports that the update did not complete and points at `update.log` (the update's own output). It never reloads blindly onto the same version. |
 | `BrowserAccessSection` / `OpenControlUiButton` | `app/agents/BrowserAccessSection.tsx` | Browser access to one agent's Control UI, in its detail panel: requests waiting for approval (Approve / Reject) and approved browsers (Rename / Revoke), refreshed every 5 s while mounted. A successful approve, reject, rename or revoke updates the list at once, since the refresh behind it runs the OpenClaw CLI and takes seconds; a read started before the mutation is discarded. On agents that require approval, "Invite link" fetches `/api/agents/[id]/invite-link` and shows the link in a read-only field with Copy, which uses the Clipboard API in a secure context and the field's selection over plain HTTP, plus a warning when the link uses localhost. `OpenControlUiButton` opens `/api/agents/[id]/open-control-ui` in a new tab inside the click; that route redirects to a one-time link that pairs the browser with no approval, or to the plain token link when none can be issued. Used on the agent cards and in the panel. |
 | `ChannelManager` / `ChannelSection` | `app/agents/ChannelManager.tsx`, `ChannelSection` inline in `app/agents/PageClient.tsx` | Telegram, in the agent detail panel (`ChannelSection` is the card that opens the modal; the modal title is the agent's display name). Reads `GET /channels`; lists pending pairing requests with **Approve** (`POST /channels/pairing`) and approved senders with **Revoke** after a confirm (`DELETE /channels/pairing?senderId=`). Pending comes from `openclaw pairing list`, approved from OpenClaw's pairing store (`lib/channelManager.ts`). When that store cannot be read the panel shows the reason instead of "No approved senders" (`PairingState.error`), so an empty list is never a guess. Polls pairings every 5 s while open; one keyed busy state per action. |