agentgui 1.0.1006 → 1.0.1008

package/AGENTS.md

@@ -1,5 +1,9 @@
 # AgentGUI — Agent Notes
 
+## GUI completion sweep (2026-07-02) — twenty-fourth run
+
+`gui-completion` workflow wf_4a731841-832 (33 agents, 17 confirmed). Landed 5 of the smaller-scope findings: Live-tab dashboard state (sort/filter/errors-only) now round-trips through the URL hash (`lsort=`/`lfilter=`/`lerr=1`) and localStorage — previously `lv.filter` wasn't persisted anywhere and the whole dashboard view reverted to defaults on reload. Shortcuts overlay (`?`) now wraps its content in the kit's existing `FocusTrap` component instead of a bare `role=dialog` div — Tab could previously escape into the page behind the overlay. History event rows expose copy even when collapsed (was expanded-only). Settings agents panel gives a not-installed direct-runner agent (claude-code/agy) a "re-check" button, matching the ACP rows' "restart" affordance. **Caught and fixed a real CSRF-guard test gap**: adding the regression test for the upload-PUT route (`scripts/validate-mutations.mjs`), the first draft accidentally included the `Authorization` header — matching every other case in the file but testing the wrong scenario (a stolen-token request, not real cross-site CSRF) — and passed trivially at 200. Corrected to omit auth (the path a real cross-site attacker is actually limited to); the server's existing guard correctly rejects it. **Rule for any CSRF/auth regression test: match the real attacker's capability set exactly — copying a nearby test's header pattern can silently test the wrong threat model.** 12 findings deferred to their own re-scoped PRD row (mid-thread retry/fork, file-management move UX with folder-tree picker, event-level permalinks, rolling context-size indicator — each needs a larger architecture change than a targeted fix). No kit changes this run. Pushed `d3e2d1e`. 28/28 tests, 17/17 validate-mutations, 0 console errors.
+
 ## GUI perceived-performance sweep (2026-07-02) — twenty-third run
 
 Small, targeted run landing the 22nd-run's deferred perceived-perf findings. `index.html` now `modulepreload`s the marked/dompurify ESM CDN modules (was preconnect-only) so the fetch starts in parallel with the kit/app module graph instead of waiting for `markdown.js`'s dynamic import on the first chat response's escaped-text-to-rendered-markdown swap. Settings server panel shows "checking…" instead of "unknown" for version/uptime/clients/projects-folder during the very first health probe (new `state.healthChecking` flag around `recheckHealth`'s await — "unknown" reads as a permanent broken-server fact, not in-flight loading). Settings agents panel title shows "agents · loading…" instead of the literal "agents · 0/0 installed" while the initial fetch is in flight. Still deferred (its own PRD row, re-scoped not dropped): first-render network-gating — the shell paints nothing between the boot-splash and health-probe completion, needing a render-architecture change (skeleton/optimistic shell) bigger than this run's slice. No kit changes this run. agentgui pushed `6a045ba`, 28/28 tests, 0 console errors.
@@ -75,19 +79,7 @@ Fluid-column + drag-resize density layer (workflow `gui-screen-realestate.js`, 5
 Visual/layout/motion parity vs claude.ai/code (workflow `gui-claude-code-parity.js`, 66 agents -> 37 gaps). Flat full-width chat turns (`.chat-msg-flat` at `--measure`), single bordered composer rounded-rect, `.chat-tool` status-toned cards, `WorkspaceShell mainFlush`, command-center `SessionDashboard`, reduced-motion-guarded transitions. **Load-bearing webjsx rule: never pass a conditional `x?h():null` positionally — build the array and `.filter(Boolean)` it; ANY null among VElement siblings crashes `applyDiff` (reading 'key'); `Btn` spreads array children.** Full detail in rs-learn (recall "agentgui 10th run parity").
 ## GUI logic+predictability sweep (2026-06-10) — ninth maximum-effort run
 
-Audits whether everything OPERATES logically and predictably (not coverage, not cohesion). New workflow `.claude/workflows/gui-logic-predictability.js` (6 lenses: interaction-lifecycle, cross-surface-consistency, information-architecture, realtime-truth, input-ergonomics, scale-robustness; hunt -> adversarial verify (kept-typography guard) -> plan). 78 agents -> 67 confirmed findings (`PUNCHLIST-LOGIC.md`). ALL implemented across kit/app/server.
-
-**Interaction lifecycles.** Kit `ConfirmDialog`/`PromptDialog` gain `error` (inside `.ds-modal-body`, role=alert) + `busy`/`busyLabel` (disables actions AND Escape/backdrop close via `data-busy`); Backdrop's Escape listener is **document-level** (focus can sit outside the dialog) and focus-restore fires only on genuine close (queueMicrotask remount guard) - the app's global Escape ladder ALSO closes `state.files.dialog` as belt-and-braces (a witness caught Escape dead with focus on H1). Per-session stop: `state.live.stopping` Set + kit `session.stopping` ('stopping...', disabled); bulk stops await `Promise.allSettled`, announce 'stopped N of M', partial failures keep their selection. Upload: one drain loop, concurrent drops append, 409 rows get `replace` (overwrite=1) + dismiss (kit `UploadProgress {actions,onDismiss}`). new-chat/backend-change/stop confirms are arm-only with bound values + 4s reset; chips never clobber a typed draft.
-
-**Realtime truth.** Dashboard joins on `realSid = claudeSessionId || sessionId` - server `chat.active` rows now carry `claudeSessionId`; `streaming_cancelled` is broadcast (an aborted turn never also completes; backend maps it to `{type:'cancelled'}`, app marks the turn 'stopped'); a 60s per-session terminal-frame buffer replays on re-subscribe (`replayedTerminal`) so a turn finishing during a ws drop cannot hang busy; `chat_active_changed` hints (started/ended/cancelled) trigger immediate refreshActive. Error status derives from recency (not cumulative count - a session is not 'error' forever); stale derivation honors any session's running tool (per-sid `toolRunning` off SSE); SSE counter increments dedup by `ev.i` maxI; counters take max(index,tally) and never move backwards; clock-skew clamps to 'just now'; external (non-owned) sessions render as read-only dashboard cards; sorts compare real timestamps with sid tiebreaker; tally pruned 24h/200-LRU. persistChat: 500ms debounce, 100-msg cap, >4KB tool payloads trimmed (`truncatedForStorage`), quota failure banners once, cross-tab `storage` listener with ts guard ('updated in another tab' banner instead of clobber).
-
-**Input ergonomics.** IME guards (composer, PromptDialog, SearchInput - `isComposing || keyCode===229` never sends). Escape ladder: shortcuts overlay > file dialog > confirmingEdit > new-chat arm > stop-arms > stop generation (and Esc in composer cancels while busy). Window-level dragover/drop preventDefault outside `.ds-dropzone` (a dropped file never navigates the page). Composer: `onPasteFiles`/`onDropFiles` kit props (image paste with no handler -> polite transient note), `.chat-composer-hint` 'Enter to send · Shift+Enter for a new line', context `bits` may carry per-bit `onClick` (cwd is its own target; agent/model inert). `Mod+Shift+L` focuses composer from anywhere (the one chord BEFORE the modifier early-return). cwd validates on blur (debounced `/api/stat` -> kit `cwdError`/`cwdChecking`). Streaming re-render pauses while a selection is held inside the thread (kit `hasSelectionInside` + autoscroll pause).
-
-**Scale.** `AgentChat {shownMessages,onShowEarlier}` windowed thread (`MESSAGE_CAP=100`, keyed 'show N earlier turns' row); >20k streaming part renders a 4k tail window ('streaming · N KB so far'); history events sliced to last 5000 in-app (server adds a `?limit=N[&before=]` slicing shim in server.js - ccsniff itself has no limit support, do NOT edit node_modules); markdown loader failure no longer latches degraded (30s retry, cache never stores escaped fallback).
-
-**Consistency/IA.** ONE `fmtDuration` (kit sessions.js) + ONE `fmtBytes` (= fmtFileSize, `0 B` not em-dash) kit-wide with app fallbacks; SessionCard `title` shares the rails' title expression; `session.external` + `session.elapsedMs`; STREAM_WORD `offline`; coarse-pointer always-visible file row actions; idle status never reads 'resuming...' (plain 'continues session <sid8>' in composer context); `/` focuses the files/live filter; running panel links 'open in live'; settings reordered backend -> server -> agents -> appearance -> keyboard -> data; crumb says 'settings' and Files crumb carries the parent trail.
-
-**Witness.** localhost:3000/gm/ 0 console/page errors across chat/files/live/history/settings; composer hint renders; rename dialog opens + Escape closes + focus restores; 64-row files grid + crumb trail; 420px chat+live no h-scroll; decorative-glyph count 0 (kept middot/ellipsis/dash). Joint risk to re-check after parallel agent runs: prop-name contracts (this run caught app `onPasteImage` vs kit `onPasteFiles`, and backend missing the new `streaming_cancelled` type).
+78-agent workflow, 67 confirmed findings, all implemented (interaction lifecycles, realtime truth, input ergonomics, scale, consistency). **Escape ladder order: shortcuts overlay > file dialog > confirmingEdit > new-chat arm > stop-arms > stop generation** (still binding - later runs extend the ladder, never reorder it). Full detail in rs-learn (recall "agentgui 9th run logic predictability sweep").
 
 ## GUI completion sweep (2026-06-10) — eighth maximum-effort run
 

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "agentgui",
-  "version": "1.0.1006",
+  "version": "1.0.1008",
   "description": "Multi-agent ACP client with real-time communication",
   "type": "module",
   "main": "electron/main.js",

package/scripts/validate-mutations.mjs

@@ -39,6 +39,18 @@ await check('mkdir-ads-colon', 400, await post('/api/mkdir', { dir: ROOT, name:
 // one). On a PASSWORD-gated server the auth wall rejects it first (401);
 // open servers hit the CSRF guard (403). Either way the mutation is refused.
 await check('csrf-cross-site-form', process.env.PASSWORD ? 401 : 403, await post('/api/delete', 'path=x', { 'Content-Type': 'application/x-www-form-urlencoded', 'Sec-Fetch-Site': 'cross-site' }));
+// The upload PUT is octet-stream, not a form-postable content-type, but the
+// CSRF guard's stated reasoning (rides the same-origin SPA fetch) only holds
+// if a cross-site PUT is actually rejected the same way POST/DELETE are.
+// Deliberately omits AUTH here (unlike every other case in this file): the
+// guard's `authed` bypass is legitimate for a Bearer token an attacker page
+// cannot read, but including it here would test the wrong thing (a stolen-
+// token scenario, not CSRF) - this asserts the cookie/no-credential path a
+// real cross-site attacker is actually limited to.
+await check('csrf-cross-site-upload', process.env.PASSWORD ? 401 : 403,
+  await fetch(BASE + '/api/upload-file?dir=' + encodeURIComponent(ROOT) + '&name=csrf-probe.txt', {
+    method: 'PUT', headers: { 'Sec-Fetch-Site': 'cross-site' }, body: 'x',
+  }).then(async r => ({ status: r.status, body: await r.text() })));
 // round trip
 await check('mkdir-ok', 200, await post('/api/mkdir', { dir: ROOT, name: '_wtest' }));
 const up = await fetch(BASE + '/api/upload-file?dir=' + encodeURIComponent(join(ROOT, '_wtest')) + '&name=up.txt', { method: 'PUT', headers: AUTH, body: 'hello upload' });

package/site/app/js/app.js

@@ -3,7 +3,7 @@ import * as B from './backend.js';
 
 installStyles().catch(() => {});
 
-const { AppShell, WorkspaceShell, WorkspaceRail, Topbar, Crumb, Side, Status, Chat, ChatComposer, AgentChat, ConversationList, SessionDashboard, Row, Panel, PageHeader, SearchInput, TextField, Select, Btn, Icon, EventList, Spinner, Alert, FileGrid, FileSkeleton, sortFiles, FileToolbar, RootsPicker, BreadcrumbPath, EmptyState, FileViewer, FilePreviewPane, FilePreviewCode, FilePreviewText, FilePreviewMedia, ThemeToggle, ContextPane, PromptDialog, ConfirmDialog, DropZone, UploadProgress, FilterPills, SessionMeta, BulkBar, Checkbox, ShortcutList } = C;
+const { AppShell, WorkspaceShell, WorkspaceRail, Topbar, Crumb, Side, Status, Chat, ChatComposer, AgentChat, ConversationList, SessionDashboard, Row, Panel, PageHeader, SearchInput, TextField, Select, Btn, Icon, EventList, Spinner, Alert, FileGrid, FileSkeleton, sortFiles, FileToolbar, RootsPicker, BreadcrumbPath, EmptyState, FileViewer, FilePreviewPane, FilePreviewCode, FilePreviewText, FilePreviewMedia, ThemeToggle, ContextPane, PromptDialog, ConfirmDialog, DropZone, UploadProgress, FilterPills, SessionMeta, BulkBar, Checkbox, ShortcutList, FocusTrap } = C;
 
 // One duration/bytes vocabulary across every surface: prefer the kit's shared
 // formatters (exported alongside the components), fall back to the local
@@ -51,7 +51,7 @@ const ARM_RESET_MS = 4000;
 
 // Full routable param set. Every view-defining piece of state round-trips
 // through the hash so reload and Back/forward restore the exact view.
-const HASH_KEYS = ['tab', 'sid', 'dir', 'file', 'q', 'project', 'section', 'filter'];
+const HASH_KEYS = ['tab', 'sid', 'dir', 'file', 'q', 'project', 'section', 'filter', 'lsort', 'lfilter', 'lerr'];
 function readHash() {
   const hash = location.hash || '';
   const out = {};
@@ -82,6 +82,12 @@ function buildHash() {
     if (state.projectFilter) parts.push('project=' + encodeURIComponent(state.projectFilter));
   }
   if (tab === 'settings' && state.settingsSection) parts.push('section=' + encodeURIComponent(state.settingsSection));
+  if (tab === 'live') {
+    const lv = state.live || {};
+    if (lv.sort && lv.sort !== 'status') parts.push('lsort=' + encodeURIComponent(lv.sort));
+    if (lv.filter) parts.push('lfilter=' + encodeURIComponent(lv.filter));
+    if (lv.errorsOnly) parts.push('lerr=1');
+  }
   return parts.length ? '#' + parts.join('&') : '';
 }
 function writeHash({ push = false } = {}) {
@@ -602,10 +608,12 @@ function view() {
 
   const shortcutsHint = state.showShortcuts
     ? h('div', { key: 'sc', role: 'dialog', 'aria-modal': 'true', 'aria-label': 'Keyboard shortcuts', class: 'ds-alert ds-alert--info shortcuts-dialog' },
-        h('div', { key: 'sc-head', class: 'ds-alert-head' },
-          h('span', { key: 'sc-title', class: 'ds-alert-title' }, 'Keyboard shortcuts'),
-          h('button', { key: 'sc-close', type: 'button', class: 'ds-btn', 'aria-label': 'Close keyboard shortcuts', ref: (el) => { if (el) requestAnimationFrame(() => el.focus()); }, onClick: () => { state.showShortcuts = false; render(); announce('shortcuts closed'); } }, 'close')),
-        h('div', { key: 'sc-body', class: 'ds-alert-body' }, ShortcutList({ shortcuts: SHORTCUTS })))
+        FocusTrap({ children: [
+          h('div', { key: 'sc-head', class: 'ds-alert-head' },
+            h('span', { key: 'sc-title', class: 'ds-alert-title' }, 'Keyboard shortcuts'),
+            h('button', { key: 'sc-close', type: 'button', class: 'ds-btn', 'aria-label': 'Close keyboard shortcuts', onClick: () => { state.showShortcuts = false; render(); announce('shortcuts closed'); } }, 'close')),
+          h('div', { key: 'sc-body', class: 'ds-alert-body' }, ShortcutList({ shortcuts: SHORTCUTS })),
+        ] }))
     : null;
   const main = h('div', { id: 'agentgui-main', role: 'region', 'aria-label': 'main content', 'data-chat-scroll': '', class: 'agentgui-main agentgui-main-' + state.tab }, [shortcutsHint, ...mainContent()].filter(Boolean));
 
@@ -1467,7 +1475,8 @@ function filesMain() {
 const LIVE_PREFS_KEY = 'agentgui.live';
 const FILES_PREFS_KEY = 'agentgui.files';
 function persistLivePrefs() {
-  lsSet(LIVE_PREFS_KEY, JSON.stringify({ sort: state.live.sort || 'status', errorsOnly: !!state.live.errorsOnly }));
+  lsSet(LIVE_PREFS_KEY, JSON.stringify({ sort: state.live.sort || 'status', errorsOnly: !!state.live.errorsOnly, filter: state.live.filter || '' }));
+  if (state.tab === 'live') writeHash();
 }
 function persistFilesPrefs() {
   lsSet(FILES_PREFS_KEY, JSON.stringify({ sort: state.files.sort || 'name', sortDir: state.files.sortDir || 'asc', density: state.files.density || 'list' }));
@@ -1475,7 +1484,7 @@ function persistFilesPrefs() {
 function hydratePrefs() {
   try {
     const lv = JSON.parse(lsGet(LIVE_PREFS_KEY) || 'null');
-    if (lv) { if (lv.sort) state.live.sort = lv.sort; state.live.errorsOnly = !!lv.errorsOnly; }
+    if (lv) { if (lv.sort) state.live.sort = lv.sort; state.live.errorsOnly = !!lv.errorsOnly; if (lv.filter) state.live.filter = lv.filter; }
   } catch {}
   try {
     const fp = JSON.parse(lsGet(FILES_PREFS_KEY) || 'null');
@@ -1700,7 +1709,7 @@ function liveMain() {
       streamState,
       activeSid: activeCard ? activeCard.sid : state.chat.resumeSid,   // W13
       sort: { value: sortKey, onChange: (v) => { state.live.sort = v; persistLivePrefs(); render(); } },
-      filter: { value: lv.filter || '', placeholder: 'Filter sessions', onInput: (v) => { state.live.filter = v; render(); } },
+      filter: { value: lv.filter || '', placeholder: 'Filter sessions', onInput: (v) => { state.live.filter = v; persistLivePrefs(); render(); } },
       errorsOnly: !!lv.errorsOnly,
       onErrorsOnly: (on) => { state.live.errorsOnly = on; persistLivePrefs(); render(); },
       selectable: true,
@@ -2730,10 +2739,13 @@ function historyMain() {
             rail,
             expanded,  // disclosure state -> kit Row sets aria-expanded
             highlight: q || undefined,
-            actions: expanded ? [{
+            // Copy is available whether or not the row is expanded - a user
+            // scanning collapsed rows for a specific payload shouldn't have
+            // to expand every row first just to copy one.
+            actions: [{
               label: 'copy', title: 'copy event',
               onClick: () => copyText(full || raw || ('(' + type + ')'), 'event copied'),
-            }] : undefined,
+            }],
             title: expanded ? (typePrefix + (text || '(' + type + ')')) : (collapsedTitle || typePrefix + '(' + type + ')'),
             detail: expanded && e.toolInput ? JSON.stringify(e.toolInput, null, 2) : undefined,
             // Guard ts: a missing/zero timestamp renders "Invalid Date" otherwise.
@@ -3211,7 +3223,11 @@ function agentsPanel() {
             onClick: usable ? () => { navTo('chat'); selectAgent(a.id); } : undefined,
             right: (acp && !acp.healthy)
               ? [Btn({ key: 'acprestart', onClick: (e) => { e.stopPropagation(); B.restartAcpAgent(state.backend, a.id).then(() => loadAgents()); }, children: 'restart' })]
-              : undefined,
+              // Direct-runner agents (claude-code/agy) have no restart, but a
+              // 'not installed' verdict can go stale if the user installs the
+              // CLI mid-session - give it the same recheck affordance ACP rows
+              // get via 'restart', instead of only static prose.
+              : (!avail ? [Btn({ key: 'agrecheck', onClick: (e) => { e.stopPropagation(); loadAgents(); }, children: 're-check' })] : undefined),
           });
         })
       // The empty array means one of three things; never let an in-flight load
@@ -3501,6 +3517,11 @@ async function init() {
     }
     if (bootTab === 'history' && hp.q) state.searchQ = hp.q;
     if (bootTab === 'history' && hp.project) state.projectFilter = hp.project;
+    if (bootTab === 'live') {
+      if (hp.lsort) state.live.sort = hp.lsort;
+      if (hp.lfilter) state.live.filter = hp.lfilter;
+      if (hp.lerr) state.live.errorsOnly = true;
+    }
     navTo(bootTab, { push: false });
     if (bootTab === 'history' && state.searchQ.trim().length >= 2) runSearch();
     if (bootTab === 'settings' && hp.section) focusSettingsSection(hp.section);