ssh2 emits multiple error events per failed connection (e.g. ECONNRESET
followed by "Connection lost before handshake"). Several code paths used
`.once("error")` which removed the listener after the first event,
leaving the second error unhandled and crashing the process via the
uncaughtException handler's re-throw.
Root cause: `runDistroDetection` ran unconditionally after connection
attempts (including failures), creating a new SSHClient to the same
unreachable host. Its `execCommand` used `.once("error")`, so the
second ssh2 error event had no listener and became an uncaught exception.
Fixes:
- execCommand: `.once("error")` → `.on("error")` with settled guard and
explicit `conn.end()` cleanup
- runDistroDetection: move into try block so it only runs after
successful connections
- portForwardingBridge: same `.once` → `.on` fix
- sftpBridge: add catch-all error listener after cleanup() removes the
pre-ready listeners
- main.cjs: suppress non-fatal SSH/network errors in uncaughtException
and unhandledRejection handlers as defense-in-depth (log to crash
bridge, do not re-throw)
Closes#452
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix: support bare IPv6 addresses in quick connect and fix IPv6 display
- Accept un-bracketed IPv6 addresses (e.g. 2607:f130::4f06) in quick
connect input. The main regex requires brackets for IPv6+port, but now
falls back to detecting bare IPv6 (2+ colons, hex-only) when the
primary pattern fails.
- Add formatHostPort() helper that wraps IPv6 addresses in brackets
when appending a port, preventing ambiguous displays like
"2607:f130::4f06:22"
- Apply formatHostPort in QuickConnectWizard, TerminalConnectionDialog,
and SftpSidePanel
- Fix hop label formatting in sshBridge and sftpBridge for IPv6 jump
hosts
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix: truncate long hostnames in connection dialog
Add truncate to the host label and protocol subtitle in the connection
dialog so long IPv6 addresses don't overflow into the action buttons.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix: constrain connection dialog header so truncate works correctly
Add min-w-0/flex-1 to the left side of the header flex container and
shrink-0 to the avatar so long hostnames truncate instead of pushing
into the Show logs / close buttons.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix: prevent action buttons from being squeezed by long hostname
Add shrink-0 and left margin to the right-side button group so truncated
text doesn't crowd into Show logs / close buttons.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix: tighten bare IPv6 detection to avoid MAC address false positives
Only accept bare (un-bracketed) hex:colon strings as IPv6 if they
contain '::' (unambiguously IPv6) or have exactly 7 colons (full
8-group notation). This rejects MAC addresses like aa:bb:cc:dd:ee:ff
(5 colons) which would otherwise trigger quick-connect mode.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix: avoid double-wrapping already-bracketed IPv6 hop labels
Add !startsWith('[') guard so hostnames that are already bracketed
(e.g. from URL-imported hosts) don't produce malformed labels like
[[2607:f130::4f06]]:22.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- Add smoothScrolling boolean to TerminalSettings (default: true)
- Wire setting to xterm.js smoothScrollDuration (120ms when on, 0 when off)
- Add toggle in terminal settings UI
- Include in sync payload and i18n strings (en, zh-CN)
Inspired by #467 (@crawt).
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- Use a regex ASCII test to detect lines where string indices equal cell
columns, skipping the buildStringToCellMap buffer walk entirely. Most
terminal output is ASCII, so this avoids the majority of cell API calls.
- Share a frozen empty array for non-matching lines instead of allocating
a new array per scanLine call, reducing GC pressure during scrollback.
Inspired by #466 (@crawt).
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* perf(keyword-highlight): reduce highlight latency with throttled rAF and line cache
Based on #464 by @crawt with fixes for review feedback:
- Split triggerRefresh into immediate (rAF) and debounced (setTimeout) modes
so onWriteParsed highlights land with fresh content instead of trailing
by 200ms
- Throttle the immediate path (50ms min interval) to prevent heavy output
like tail -f from refreshing every frame
- Add per-line match result cache (LRU, bounded by cacheEntries config)
so repeated or scrolled-back lines skip regex scanning entirely
- Lazily build cellMap only when a regex match is found, avoiding
unnecessary work on non-matching lines
- Fix buildStringToCellMap to handle empty cells (codepoint 0) which
translateToString() renders as spaces — keeps the map aligned with
the string and makes lineText a safe cache key
- Clean up animationFrameId and matchCache on dispose/rule change
Co-Authored-By: Leo Pan <crawt@users.noreply.github.com>
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix: guard rAF callback against stale state and add debounce fallback
- Re-check enabled/alternate-buffer inside the rAF callback so a
pending frame doesn't resurrect decorations after the user disables
highlighting or enters an alternate-buffer app
- Schedule a debounce timer alongside rAF so background/hidden tabs
(where Chromium suspends rAF) still get highlight updates
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix: prevent fallback timer from being cleared on rAF-pending path
- Don't clear debounceTimer at the start of immediate mode — in hidden
tabs rAF stays pending indefinitely, so repeated onWriteParsed calls
were clearing the only timer that could actually fire
- Cancel debounceTimer inside the rAF callback instead, so foreground
tabs don't get a redundant second refreshViewport() 200ms later
- Only arm a new fallback timer if one isn't already pending
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix: clear stale rAF in fallback timer and add alternate buffer guard
- Cancel the pending rAF and clear animationFrameId in the fallback
timer callback so hidden-tab refreshes don't leave animationFrameId
stuck, which would block all future immediate refreshes
- Add enabled/alternate-buffer re-check in the fallback callback,
matching the guard already present in the rAF callback
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix: extract executeRefresh to ensure all timer paths clear stale rAF
A debounced-path timer (from scroll/resize) could fire without clearing
a stale animationFrameId left by an earlier immediate-path rAF that
never executed (hidden tab). This left the immediate path permanently
blocked.
Extract executeRefresh() with rAF cleanup + state guards, used by all
three callback sites (rAF, immediate fallback, debounced timer).
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Leo Pan <crawt@users.noreply.github.com>
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* perf(keyword-highlight): reduce highlight latency and redundant regex scanning
- Split triggerRefresh into two modes: "immediate" (rAF, for new output
and rule changes) and "debounced" (setTimeout, for scroll/resize),
eliminating the fixed 200ms delay after each write that caused visible
highlight lag on commands like `ls`.
- Add per-line match result cache (LRU, bounded by cacheEntries config)
so repeated or scrolled-back lines skip regex scanning entirely.
- Lazily build the string-to-cell column map only when a regex match is
actually found, avoiding unnecessary work on non-matching lines.
- Clean up animationFrameId and matchCache on dispose/rule change to
prevent leaks and stale results.
* fix: include cell layout in highlight cache key to prevent misplaced decorations
Two IBufferLines can produce identical translateToString() output but
differ in cell layout (e.g. empty cells vs real space characters after
tab stops). Using lineText alone as the cache key could return cached
x/width ranges computed from a different cell layout, producing
misplaced or truncated highlights.
Build the cellMap eagerly and include it in the cache key so lines with
different cell structures get separate cache entries. Pass the pre-built
cellMap into scanLine to avoid redundant work.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: panwk <panwukan@suangoo.com>
Co-authored-by: bincxz <16399091+binaricat@users.noreply.github.com>
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix: resolve SSH chain connection hang and improve connection progress
- Fix Promise never settling when conn 'close' fires before 'ready'
during chain connections, which caused "reply was never sent" error
- Replace fake timed progress animation with real backend events
- Send granular connection progress for all SSH connections (not just
chain), including: connecting, key exchange, auth attempts, forwarding,
shell opening
- Surface auth method attempts (SSH agent, key names, password) in
progress logs so users can diagnose authentication failures
- Include error details in progress events for better error visibility
Closes#463
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix: scope progress events by sessionId, prevent duplicate errors, hide chain UI for direct SSH
- Add sessionId to chain progress payload so events are scoped per session (P1)
- Set settled=true in error/timeout handlers to prevent close handler from
emitting a second misleading 'closed unexpectedly' error (P2)
- Only show chain progress UI when total > 1 so direct SSH connections
don't render as 'Chain 1/1' (P3)
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix: mark shell-open failure as settled before closing connection
The conn.shell() error branch calls conn.end() which triggers the close
handler, but settled was not set yet, causing a duplicate 'closed
unexpectedly' error to overwrite the real shell-open failure message.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- TerminalLayer: remove bracket paste wrapping since we can't check
term.modes.bracketedPasteMode here — keep only normalizeLineEndings
- createXTermRuntime: broadcast un-wrapped data before applying
bracket paste, so target sessions don't receive literal escape
sequences meant for the source terminal's paste mode state
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Only wrap multi-line snippets in bracket paste sequences when:
- createXTermRuntime: term.modes.bracketedPasteMode is active AND
disableBracketedPaste setting is false (matches paste handler)
- TerminalLayer: disableBracketedPaste setting is false (no access
to term.modes, but respects user opt-out)
Prevents sending literal ^[[200~ escape sequences to shells that
don't support or have disabled bracketed paste mode.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Snippet execution via sidebar click was missing normalizeLineEndings()
and bracket paste wrapping that the paste handler and shortkey handler
already apply. On Windows ConPTY/PowerShell, sending raw multi-line
input without bracket paste can cause out-of-order line execution
because the shell processes lines individually and asynchronously.
- Add normalizeLineEndings() to sidebar snippet click handler
- Wrap multi-line snippets in bracketed paste sequences (\e[200~...\e[201~)
so the shell treats them as a single atomic paste
- Apply same fix to shortkey snippet handler for consistency
- Fix broadcast payload to use the processed data
Fixes#455
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Replace flex-wrap layout with single-line truncate + title tooltip
for the environment metadata row, preventing awkward wrapping when
the settings window is narrow.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- Extract error properties (code, errno, syscall, hostname, port,
signal, level) into errorMeta field for system-level diagnostics.
- Add extra field for structured context (e.g. render-process-gone
reason and exitCode as separate fields, not just a string).
- Add process PID for correlating with OS-level logs.
- Accept optional extra parameter in captureError() for callers to
attach structured context data.
- Display errorMeta and extra as tagged badges in the crash log viewer.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Add electronVersion, osVersion, memoryUsage (RSS/heap in MB),
activeSessionCount, and process uptime to each crash log entry.
Display these fields inline in the Settings crash log viewer.
These extra fields help diagnose issues like #452 where knowing
the session count and memory state at crash time is critical.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- Mark re-thrown unhandledRejection errors so uncaughtException handler
skips duplicate logging.
- Reload crash log list after clearing instead of blindly emptying,
so partial delete failures still show remaining files.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- Move crashLogBridge require before process error handlers so it is
available if a bridge import throws during startup.
- Add request ID ref to handleExpandCrashLog to discard out-of-order
results when the user clicks different log files in quick succession.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Capture main-process errors (uncaughtException, unhandledRejection,
render-process-gone) to JSONL log files in userData/crash-logs/ with
30-day auto-rotation. Users can view, expand, and clear crash logs
from Settings > System to help diagnose issues like #452.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Add 'sftp.transfer.preparing' key to en.ts and zh-CN.ts so the
indeterminate transfer state shows localized text instead of the
raw i18n key.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* chore: remove 65 debug console.log statements from production code
Remove bracketed debug traces ([SFTP navigateTo], [SFTPBackend],
[ManagedSourceSync], [AutoSync], [CloudSync], [Settings], etc.)
across 16 files. These were development logging that shipped to
production, creating noise in the console.
Also clean up dead variables left behind after log removal
(hotkeyDebug, results, verification reads).
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* chore: remove 43 unused exports and dead type definitions
Remove export keywords from symbols that are never imported outside
their defining file. Symbols still used internally keep their
definitions; symbols not used at all are removed entirely.
Removed entirely: TerminalLine, SessionLogsSettings, KDFParams,
SyncManagerConfig, GoogleTokenResponse, OneDriveTokenResponse,
getSyncStatusColor, resolveHostTerminalAppearance,
TerminalAppearanceDefaults.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
When creating a new host, the global fontSize and theme were copied
into the host config. Since fontSizeOverride/themeOverride were not
set (undefined), the legacy detection logic treated the presence of
these values as an active override, locking the host to the global
values at creation time.
Stop copying fontSize and theme into new host configs. Without these
fields, resolveHostTerminalFontSize/ThemeId correctly falls back to
the current global setting, so hosts dynamically follow global
changes unless the user explicitly sets a per-host override.
Closes#424
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix: resolve SFTP tab connection key race condition in workspace mode
When rapidly switching focus between workspace panes, the single
pendingConnectionKeyRef could be overwritten before the tracking
effect mapped it to the created tab. This left tabs unmapped in
tabConnectionKeyMapRef, causing duplicate tabs on subsequent switches.
Replace the two-step async mechanism (pendingConnectionKeyRef + deferred
tracking effect) with a synchronous onTabCreated callback on connect().
The callback fires immediately after the tab ID is determined, before
any async SSH work begins, eliminating the race window entirely.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix: scope SFTP transfers to active connection and prevent stale session lookups
Two fixes for workspace focus-switching issues:
1. Transfer queue now filters by the active connection's host, so
switching focus between workspace panes only shows transfers
relevant to the currently displayed SFTP tab.
2. Move sftpSessionsRef.delete() before the async closeSftp() call
to close the race window where concurrent code could look up a
stale sftpId that the backend has already removed, causing
"SFTP session not found" errors.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix: allow SFTP focus switching during file transfers
Active transfers should not block workspace focus-following. Transfers
run on their own sftpId independent of the active tab, and forceNewTab
preserves old connections, so switching focus is safe.
Only interactive operations (text editor, permissions dialog, file
opener, file watches) still block host switching.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix: refresh correct SFTP tab after transfer completes during focus switch
When a transfer completes while focus has switched to a different host,
refresh was targeting the currently active pane instead of the pane that
initiated the transfer.
Add optional tabId parameter to navigateTo() and refresh() so callers
can target a specific tab. Capture the tab ID at transfer start and use
it for the post-transfer refresh, ensuring the correct tab's file list
is updated regardless of which tab is currently focused.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix: auto-reconnect SFTP when session is lost during navigation
When navigateTo() detected a missing or expired SFTP session, it
cleared the connection to null, showing the empty "Select a host"
state. Now it delegates to handleSessionError(), which triggers the
existing reconnection mechanism — keeping files visible while
reconnecting in the background.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* perf: eliminate redundant stat calls before file transfers
Before this change, each file transfer performed 3-4 stat calls over
the network before the progress bar started moving:
1. startTransfer: stat to get file size (~100ms)
2. processTransfer: stat again if size was 0 (~100ms)
3. Conflict check: stat source file for mtime (~100ms)
4. Backend: stat again if totalBytes missing (~100ms)
Now:
- Use the source pane's cached file list for size and mtime (zero
network cost) instead of stat calls in startTransfer
- Store sourceLastModified on TransferTask so the conflict check can
use it directly instead of a redundant source stat
- Backend already skips stat when totalBytes is provided
This saves ~200-300ms of network round-trips per file before the
progress bar starts moving.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* perf: show immediate progress feedback during transfer setup
The progress bar previously stayed at 0% for ~500ms-1s while the
backend acquired an isolated SFTP channel and waited for the first
data chunk. Users perceived this as the transfer being "sluggish".
Now start simulated progress immediately for all single-file
transfers (not just non-streaming ones). When the first real progress
update arrives from the backend, the simulation is stopped and real
progress takes over seamlessly. This gives instant visual feedback
that the transfer is in progress.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix: show accurate transfer progress instead of simulated values
The progress system had fundamental issues:
1. Simulated progress ran for ALL transfers including streaming ones,
creating fake progress that could reach 95% while real progress
was at 60%. The Math.max ratchet prevented regression, so users
saw inflated numbers.
2. Speed and remaining time were based on simulated data during the
setup phase, giving misleading estimates.
Changes:
- Only use simulated progress for non-streaming transfers (no real
progress callback available). Streaming transfers get real data.
- Remove the double ratchet (Math.max) from onProgress — the backend
already enforces monotonic progress, so the frontend should trust
the reported values directly.
- Show an indeterminate "preparing..." state during the setup phase
(channel acquisition, conflict check) instead of fake progress.
This honestly communicates that the transfer is starting.
- Hide speed and remaining time during the indeterminate phase since
no real data is available yet.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* refactor: remove dead progress simulation and non-streaming transfer code
startStreamTransfer is always available in Electron, so:
- Remove the non-streaming fallback path in transferFile() that read
entire files into memory with no progress reporting
- Remove startProgressSimulation / stopProgressSimulation and all
related refs (progressIntervalsRef, useSimulatedProgress,
hasStreamingTransfer)
- Remove the cleanup effect for progress intervals
All transfers now use the streaming path with real backend-reported
progress. The indeterminate "preparing..." state covers the setup
phase until the first real progress arrives.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* perf: reduce SFTP transfer concurrency from 64 to 4
64 parallel SFTP read/write requests overwhelmed servers, causing
the first chunk response to be delayed by 46+ seconds. Reducing to
4 concurrent requests provides a responsive first progress update
(~1-2s) while still offering significant speedup over sequential
streaming.
Also adds timing logs to the transfer pipeline (processTransfer,
transferFile, downloadFile, uploadFile) to aid future diagnostics.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix: address review findings from PR #443
Critical fixes:
- Fix refresh/navigateTo type signatures to include the tabId option
parameter — previously it was silently ignored, making tab-targeted
refresh non-functional
- Fix handleSessionError/reconnection in navigateTo for background tabs:
when called with explicit tabId, update that specific tab instead of
the active tab (which could be a different host)
- Fix uploadExternalFiles to capture and pass tabId for post-upload
refresh (was missing, only uploadExternalEntries was fixed)
Medium fixes:
- Restore Math.max monotonic ratchet on single-file onProgress to guard
against any non-monotonic backend values
- Add stat fallback in processTransfer to populate sourceLastModified
when file is not in the pane's visible file list (filtered/search)
- Adjust TRANSFER_CONCURRENCY from 4 to 8 as a better throughput/
responsiveness balance
Cleanup:
- Remove all debug timing logs (console.log with Transfer/downloadFile/
uploadFile prefixes) from both frontend and backend
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix: prevent background tab navigation from rolling back active tab
Two P1 fixes from automated review:
1. navSeqRef race: navigateTo uses a per-side sequence counter, so a
background tab refresh would bump it and cause the active tab's
concurrent navigation to think it was superseded, restoring
previousPath instead of applying the fetched files. Now when
navSeqRef is superseded but tabNavSeqRef still matches, the fetched
result is applied (it's valid for this tab — only a different tab
bumped the counter).
2. Auto-follow tear down: needsNewTab only checked hostId, so same
host with different session-time overrides (port/protocol) would
reuse the tab and close the old SFTP session, aborting any
in-flight transfer. Now needsNewTab is true whenever the current
connection is alive, always preserving it with forceNewTab.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
When switching terminal tabs, the SFTP side panel would reset to the
initial directory (terminal cwd at open time), discarding user navigation.
Root cause: an effect cleared the initialLocation guard on every
visibility transition (isVisible false→true), causing the initialLocation
effect to re-navigate to the original path. Tab switches toggle
visibility, so every tab switch triggered the reset.
Remove the visibility-based guard reset. When the panel is truly closed,
the component unmounts and refs reset naturally. Tab switches only
hide/show the panel and should preserve navigation state.
Closes#440
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Use Unicode property escapes (\p{L}, \p{N}) in validation regex so
Chinese and other non-ASCII characters are accepted when creating or
renaming snippet packages. Remove the HTML pattern attribute that
doesn't support the Unicode flag.
Closes#434
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix: log file name and use local time
* fix: improve SSH txt log sanitization with ANSI/OSC
* fix: log file name and use local time(update)
---------
Co-authored-by: yuzifu <yuzifu@TB16PGen5.Info>
* fix: respect global terminal appearance settings
* feat: add reset to global terminal appearance
* fix: preserve legacy host appearance overrides
* fix: show legacy appearance reset controls
* refactor: reorder terminal global reset actions
* refactor: present global theme as theme option
* refactor: present global font as font option
* feat: inline approval gate for tool execution
Replace SDK-level needsApproval with Promise-based approval gate inside
tool execute functions. The SDK stream stays alive while the UI shows
inline approve/reject buttons on ToolCall blocks.
Changes:
- Add approvalGate.ts: Promise-based approval system with event listeners
- tools.ts: requestApproval() inside execute for confirm mode
- tool-call.tsx: inline approval buttons and keyboard shortcuts
- ChatMessageList.tsx: subscribe to approval events, render approval UI
- useAIChatStreaming.ts: remove old useToolApproval hook integration
- AIChatSidePanel.tsx: remove old approval hook, clean up unused destructuring
- systemPrompt.ts: update confirm mode to not ask for text confirmation
- preload.cjs: filter pager env var prefixes from terminal display
- mcpServerBridge.cjs: add approval gate for ACP/MCP write operations
- aiBridge.cjs: wire IPC for MCP approval response and main window getter
- preload.cjs: add onMcpApprovalRequest/respondMcpApproval APIs
* fix: scope approval gate by chatSessionId and replay for late subscribers
Address Codex PR review comments:
- Add chatSessionId to ApprovalRequest for session isolation
- Scope clearAllPendingApprovals(chatSessionId?) to only clear
approvals belonging to the target session
- Add replayPendingApprovals() so late-mounting ChatMessageList
picks up approvals that fired while unmounted
- Scope MCP clearPendingApprovals in aiBridge cancel handler to
effectiveChatSessionId instead of clearing all
- Pass chatSessionId through MCP approval IPC flow
* chore: remove old approval flow code
- Delete useToolApproval.ts (unused hook)
- Delete InlineApprovalCard.tsx (replaced by ToolCall inline buttons)
- Remove stale comments referencing old hook in AIChatSidePanel
- Remove unused ai.chat.toolApprovalTitle i18n key from en/zh-CN
* fix: session-scoped approval gate and MCP replay survival
- handleStop passes activeSessionId to clearAllPendingApprovals
- setupMcpApprovalBridge stores MCP approvals in pendingApprovals map
so they survive ChatMessageList unmount/remount cycles
- ChatMessageList accepts activeSessionId prop and filters standalone
MCP approval blocks to the current session only
- AIChatSidePanel passes activeSessionId to ChatMessageList
* fix: filter PTY exec marker echoes and exit code lines from terminal
Extend filterMcpMarkers in preload.cjs to strip all shell-visible
artifacts from AI command execution:
- Echoed printf start marker: printf '%s\n' '__NCMCP_..._S'
- Echoed exit code restoration: (exit $__nc)
- PowerShell: Write-Output, $global:LASTEXITCODE, $__nc assignment
- Fish: set __nc $status
- Cmd: echo __NCMCP_...
- Widen guard to also trigger on __nc and PAGER=cat strings
* fix: scope SDK approvals, deny MCP on no renderer, fix memo comparator
- createCattyTools accepts chatSessionId and passes it to
requestApproval so SDK approvals can be matched by
clearAllPendingApprovals(activeSessionId) on stop
- useAIChatStreaming passes sessionId to createCattyTools
- mcpServerBridge: deny (resolve false) when no renderer window is
available instead of auto-approving, preserving confirm mode safety
- ChatMessageList: add activeSessionId to React.memo comparator so
switching sessions triggers re-render for correct MCP approval filter
* fix: MCP listener lifecycle, approval timeout, and UI sync on stop
- Move setupMcpApprovalBridge from ChatMessageList to AIChatSidePanel
so the IPC listener survives tab/panel switches
- Add 5-minute auto-deny timeout to requestApproval to prevent
indefinite isStreaming hangs when user walks away
- Add onApprovalCleared listener system: clearAllPendingApprovals now
notifies UI subscribers so ChatMessageList removes stale cards
- ChatMessageList subscribes to onApprovalCleared to sync local state
* fix: main-process approval timeout and full tool args in payload
- Add 5-minute auto-deny timeout to requestApprovalFromRenderer
matching the renderer-side requestApproval behavior
- Forward all tool params (excluding chatSessionId) to approval UI
instead of cherry-picking command/input/path, so sftpRename
oldPath/newPath and other tool-specific args are visible
* fix: move MCP bridge to TerminalLayer, narrow terminal filter guard
- Move setupMcpApprovalBridge from AIChatSidePanel to TerminalLayer
so the IPC listener stays alive regardless of side panel tab.
AIChatSidePanel only mounts when activeSidePanelTab==='ai'.
- Narrow preload.cjs filter guard back to __NCMCP_ only, preventing
false-positive stripping of user scripts containing __nc or PAGER=cat
* fix: eliminate PTY wrapper echo leakage and duplicate prompts
- Posix wrapper now emits 2 lines instead of 4: start marker + command
on line 1 (joined with ;), end marker + exit on line 2. This
eliminates the duplicate prompt echo from the separate start marker.
- Rename __nc to __NCMCP_rc in all shell variants (posix/fish/powershell)
so every wrapper variable contains the __NCMCP_ prefix. The preload
guard `data.includes("__NCMCP_")` now reliably catches ALL wrapper
artifacts regardless of chunk boundaries.
- Update all filterMcpMarkers regex patterns to match the restructured
wrapper format and renamed variable.
* fix: sync main-process approval timeout with renderer UI cleanup
- When requestApprovalFromRenderer times out, send IPC event
netcatty:ai:mcp:approval-cleared to renderer so stale approval
cards are removed
- Add onMcpApprovalCleared preload bridge for the new IPC channel
- setupMcpApprovalBridge now subscribes to cleared events, removes
timed-out entries from pendingApprovals and notifies clearedListeners
so ChatMessageList drops the stale card
* fix: surface denied inline approvals as errors in UI
- Detect error or denial payloads ("error" string or "ok: false")
returned by tools when the user denies an execution
- Set isError: true on the tool-result message so the ToolCall UI
renders it as a failure (red/rejected) instead of a success (green)
* Add AI support for local terminal sessions
* Fix local AI session metadata and shell safety
* Fix local session cloning and multi-exec errors
* Refactor local shell detection helpers
* Fix local shell helper import path
* Fix CJS imports in renderer
* Use ESM local shell helpers in renderer
* Normalize local shell paths and platform metadata
StickToBottom was configured with initial="smooth", causing a visible
elastic scroll animation every time the chat panel remounted on tab
switch. Change to initial="instant" so the scroll position snaps
immediately without animation. Streaming and resize still use smooth.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Scroll up to zoom in, scroll down to zoom out (10% per tick, range
25%-200%). Uses zoomRef to avoid stale closures so wheel + drag
always read the latest zoom level.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
If pointerup fires outside the window, dragStart was never cleared
and the image kept following the cursor. Now:
- Check e.buttons in pointermove to bail if primary button released
- Handle onPointerCancel and onLostPointerCapture to end drag
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Reset was disabled when zoom was 100%, so dragging without zooming
left no way to restore position. Track drag state separately and
keep reset enabled whenever the image has been dragged.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Add localized aria-label to reset, zoom in, zoom out, and close
buttons. Add i18n keys for common.reset, common.zoomIn, common.zoomOut
in en and zh-CN locales.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Drag position was stored in React state, causing the entire message
list to rerender on every pointermove frame. Move drag tracking to
refs and update the img transform directly via DOM, so only zoom
button clicks trigger React rerenders.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Drop the dark blurred overlay in favor of a shadow-2xl box-shadow
so the window boundary is clear without obscuring the background.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- Remove zoom > 100 restriction on drag — image can be panned at any
zoom level
- Add reset button (rotate-ccw icon) left of zoom controls with a
separator, resets zoom to 100% and position to center
- Reset button is disabled when already at default state
- Cursor shows grab at all times in the image area
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Replace width-based zoom with transform: scale() which is GPU-
accelerated and produces smooth 0.25s ease transitions when clicking
zoom in/out buttons. Drag translation is adjusted for current scale.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- Put filename, zoom controls, and close button in a single flex row
so they are properly aligned
- Add smooth animation on zoom (width 0.2s ease, transform 0.15s ease)
- Add drag-to-pan when zoomed beyond 100% (pointer capture based)
- Set min-width/min-height on modal to prevent extreme aspect ratios
from making the dialog too narrow or too short
- Container uses overflow hidden + fixed height to contain the image
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- Add zoom in/out buttons with percentage display in the title bar
- Zoom range: 25% - 200%, step 25%, resets to 100% on open
- Constrain modal max size to 800x700px to prevent oversized previews
- Scrollable image area when zoomed beyond container
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* feat: add click-to-preview for images in AI chat
Uploaded images in AI chat messages can now be clicked to open a
full-size lightbox preview. Clicking the overlay or the image again
dismisses it. Uses the existing Radix Dialog component.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix: use standard dialog style for image preview with close button
Replace transparent borderless overlay with proper windowed dialog that
has a background, border, and the built-in close button (X) in the
top-right corner. Remove focus ring that caused the blue border.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix: add title bar with filename and blurred backdrop to image preview
- Show filename in dialog header with border separator
- Add overlayClassName prop to DialogContent for per-instance overlay
customization (e.g. backdrop blur, custom background)
- Apply semi-transparent black background with backdrop-blur on overlay
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix: align title and close button vertically in image preview
Adjust header padding and close button position so the filename and
X button sit on the same visual line.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix: bundle claude-code-acp to prevent crash when binary is missing (#400)
When users select Claude Code in the AI module, the app spawns
`claude-code-acp` via ACP. Previously only the `claude` CLI was checked
during agent discovery, so if `claude-code-acp` was not on PATH the
spawn would fail with ENOENT and crash the Electron main process.
- Add `@zed-industries/claude-code-acp` as a bundled dependency
- Add `resolveClaudeAcpBinaryPath()` that checks PATH first, then
falls back to the npm-bundled binary (mirrors Codex pattern)
- Use the resolver in both the primary and fallback ACP provider paths
- Update agent discovery to detect agents via bundled ACP binary when
the standalone CLI is not installed
Closes#400
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix: add claude-code-acp and its deps to asarUnpack
In packaged Electron builds, files inside app.asar cannot be executed
by child_process.spawn. Add claude-code-acp and its runtime dependencies
to asarUnpack so the binary is accessible in production.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix: migrate from deprecated claude-code-acp to claude-agent-acp
The @zed-industries/claude-code-acp package has been renamed to
@zed-industries/claude-agent-acp (bin: claude-agent-acp). Update all
references across the codebase:
- package.json: replace dep with @zed-industries/claude-agent-acp@0.22.2
- electron-builder.config.cjs: update asarUnpack entries, remove stale
deps (diff, minimatch) no longer needed by the new package
- shellUtils.cjs: update binary name and require.resolve path
- aiBridge.cjs: update acpCommand, ALLOWED_AGENT_COMMANDS, isClaudeAgent
- settings types, i18n locales: update command references
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- Remove unused sessionLog deps from useCallback in App.tsx
- Wrap countAllHostsInNode in useCallback and add to useMemo deps
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix: implement real-time session logging via main process streams
Fixes#394. Session logs previously only captured ~55 lines (the
xterm serialize buffer) and were written only on session close. This
change intercepts terminal data in the main process and writes it to
a file stream in real-time, capturing the complete session output.
- Add sessionLogStreamManager.cjs: manages per-session write streams
with 500ms/64KB flush, supports txt/raw/html formats
- sshBridge: start stream on shell open, append on data, stop on close
- terminalBridge: same for local, telnet, mosh, serial sessions
- Thread sessionLog config from renderer settings through IPC options
- Skip old renderer-side auto-save when streaming is active
- Cleanup all streams on app quit
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix: remove stale renderer-side auto-save and async HTML finalization
- Remove dead renderer-side auto-save code (main process handles it)
- Make stopStream async, await writeStream finish before HTML conversion
- Use fs.promises for HTML read/write/unlink
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>