When the AI sidebar is dragged narrow, the model chip used to wrap
to its own line (and the perm chip to a third line), wasting vertical
space. Switch the toolbar from flex-wrap to a single-row flex container
where the +/perm chips stay shrink-0 and the model chip absorbs all
the squeeze via min-w-0, letting the existing truncate kick in and
ellipsize the model label instead of wrapping.
Co-authored-by: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
* feat(ai): provider switcher chip in the Catty Agent chat input
Catty Agent currently has no model chip in the chat input — modelPresets
is empty for the catty agentId so `hasModelPicker` is false. The
provider/model the chat ends up using comes from the global
`activeProviderId` / `activeModelId`, which means switching providers
requires opening Settings → AI → Providers. Closes the gap surfaced in
the original feedback on #1101 (problem 2) and partially addresses
#986 (per-context model switching) by going per-agent.
State layer adds an `agentProviderMap` (`Record<agentId, providerId>`)
alongside the existing `agentModelMap`. Both are written together when
the user picks from the new dropdown; together they form a per-agent
override that beats the global active provider/model when set. Cross-
window sync mirrors the agentModelMap pattern.
ChatInput grows a `providerSwitcher` prop carrying the enabled
ProviderConfigs, the selected (providerId, modelId), and an onSelect
callback. When supplied, the model chip switches from the Cpu glyph
to the provider's ProviderIconBadge + `providerName · modelId`, and
the popover renders a two-column layout — providers on the left
(icon + name + default model caption), that provider's known model(s)
on the right. ACP agents (Claude/Codex) are untouched because they
plumb their provider through the CLI, not the Vercel AI SDK.
AIChatSidePanel resolves `effectiveActiveProvider` /
`effectiveActiveModelId` for the catty agent from the per-agent map,
falling back to the global selection. handleSend now passes those
through to sendToCattyAgent so a provider picked in one Catty session
applies everywhere Catty runs.
For v1, each provider's right column shows just its configured
`defaultModel`. The two-level structure is in place to absorb richer
listings (cached /models, manual additions) later without UI
surgery — left a note in the popover code.
Refs #1101, #986.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
* polish(ai): flatten provider picker to a single list
Two-level dropdown was theatre — each ProviderConfig exposes exactly
one model (its `defaultModel`), so the right column ended up showing
the same string already captioned on the left row. Picking a provider
implicitly picks its model; one click is enough.
Also drops the `p.enabled !== false` filter on the picker list. The
user's expectation, confirmed by feedback on the first cut, is that
the chat-input list mirrors what Settings → AI → Providers shows. The
per-provider `enabled` toggle is an "active-ish" flag, not a
visibility gate; hiding disabled providers in the picker silently
made everything but the one currently active disappear, which is what
the reviewer hit.
Pares the popover down to a vertical list — provider icon + name +
default-model caption (mono if set, italic "configure default model"
hint if not) + a Check on the bound row. Removes the `hoveredProviderId`
state along with the right-column rendering since it's no longer
driving anything.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
* polish(ai): tune provider icon sizes in the chat input
Chip icon was sitting at sm (20px badge) and crowded the h-6 toolbar
chip alongside the truncated label; popover rows used the same sm, so
nothing differentiated the picker from the chip visually.
Adds an xs size to ProviderIconBadge (16px badge, 10px glyph, 9px
letter fallback) and uses it on the chip. Popover rows step up to md
(32px badge, 16px glyph) and the row padding grows accordingly so the
larger brand mark has room to breathe — the picker now reads like a
list of choices, the chip reads like a status line.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
* fix(ai): keep per-agent override honest in model fallback + chip label
Two codex P1/P2 findings on #1107:
P1 — When Catty has a per-agent provider override but no per-agent
model (agentModelMap['catty'] empty), cattyAgentModelId fell through
to the global `activeModelId`. That id belongs to whichever provider
was globally active, not the one Catty is now bound to, so e.g.
overriding to DeepSeek without a defaultModel happily sent gpt-4o
(global) to DeepSeek and produced a wrong-model error. The fallback
is now: stored agent model → override provider's defaultModel → empty.
Only the no-override path keeps the activeModelId tail; the
no-provider send guard catches the empty case for everyone else.
P2 — ChatInput's selectedSwitcherProvider used `?? providers[0]` so
the chip always displayed *some* provider even when none was actually
bound (selectedProviderId missing). The rest of the pipeline (send
guard, agentProviderMap) treats that as "no provider", so the chip
was lying. Dropped the fallback; when nothing is bound the chip now
shows a generic Cpu glyph + "Select provider" label until the user
picks one from the popover. Adds `ai.chat.selectProvider` in en /
zh-CN / ru.
Refs codex review on #1107.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
* fix(ai): purge per-agent binding when the bound provider is deleted
Codex local review (round 2) flagged that a saved Catty model id could
outlive its provider. The chain:
- catty bound to provider A (DeepSeek): agentProviderMap['catty']='A',
agentModelMap['catty']='deepseek-v4-flash'
- user deletes provider A
- cattyAgentProvider falls back to the global active provider (B, OpenAI)
- cattyAgentModelId still returned the saved 'deepseek-v4-flash'
- send dispatched the DeepSeek model id to OpenAI → wrong-model error
Two layers of fix:
1. **Defensive resolution** (AIChatSidePanel) — cattyAgentModelId now
looks at the override provider before trusting the stored model id.
If the override is stale (provider deleted), the stored model id is
treated as orphan and the resolution falls back to the global active
selection just like cattyAgentProvider already does.
2. **Cleanup on remove** (useAIState.removeProvider) — when a provider
is deleted, any agent whose agentProviderMap entry pointed at it
gets both maps cleared (provider override + saved model). Same
rationale: that model id is now meaningless against every other
provider. Mirrors the existing activeProviderId cleanup.
Adds a small agentProviderMapRef so removeProvider can snapshot which
agents to clean up without relying on the captured-at-callback-create
state (which would be stale).
Refs codex local review on #1107.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
* fix(ai): include agentProviderMap in settings sync + block empty-model send
Codex local review (round 3) found two issues:
P2 — `agentProviderMap` was missing from the cross-device sync surface.
Added it to the sync key list, the build path, the apply path, and the
SyncPayload type alongside the existing `agentModelMap`. Sync tests in
`syncPayload.test.ts` now cover the new field on both the build and
apply sides.
P2 — Provider rows with no `defaultModel` were still clickable in the
chat-input picker, so selecting one would save a binding with empty
model id; the send path then dispatched an empty model name and the
SDK would surface a vague backend error. Two changes:
1. ChatInput disables the row (\`disabled\`, \`aria-disabled\`, dimmed
styling, tooltip pointing the user to Settings) when the provider
has no \`defaultModel\`. Click is suppressed.
2. AIChatSidePanel \`handleSend\` adds a model-required guard mirroring
the existing no-provider guard — surfaces \`ai.chat.noProviderModel\`
as an assistant message. Catches stale bindings (e.g. user edited
the provider's \`defaultModel\` to empty after binding) before they
reach the SDK.
Refs codex local review on #1107.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
* fix(ai): trim whitespace-only model ids + reconcile orphan bindings on sync
Codex local review (round 4) found two more leaks:
P2 — Whitespace-only `defaultModel` passed the picker's `disabled` gate
(which trims) but every other layer (cattyAgentModelId resolution,
the send-guard `!sendActiveModelId` check, the SDK call) compared the
raw string. A provider whose default model was " " could still reach
the SDK and surface a vague backend error. Normalized: trim at the
resolution boundary (catty model fallback chain) and trim before the
send-guard's existence check.
P2 — Sync apply did not reconcile per-agent bindings against the
incoming provider set. A payload could change `providers` without
shipping a fresh `agentProviderMap`, leaving local overrides pointing
to provider ids the synced set no longer includes — the same ghost-
binding bug that `removeProvider` already handles for explicit user
deletes. Added `pruneOrphanPerAgentBindings()` that runs after every
AI settings apply: it walks `agentProviderMap`, drops any entry whose
provider id isn't in the current `providers` list, and clears the
saved model id for those agents alongside it (mirroring removeProvider
cleanup). A new test in `syncPayload.test.ts` exercises the ghost-
binding case end-to-end.
Refs codex local review on #1107.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
* fix(ai): notify open AI state of cross-device sync apply in the same window
Codex local review (round 4) flagged that `applySyncPayload` writes
straight to localStorage but `useAIState` only re-reads on browser
`storage` events. Those events only fire in *other* windows — the
window doing the apply (the one with the active chat panel) keeps
showing pre-sync providers and per-agent bindings until reload.
Concrete leak: cloud sync swaps in a new provider set, pruning runs,
localStorage is correct, but the open Catty chip still references the
ghost provider binding until the user reloads.
Extracts the existing `AI_STATE_CHANGED_EVENT` constant +
`emitAIStateChanged` helper out of `useAIState` and into a new
`application/state/aiStateEvents.ts` so non-React call sites (sync
apply, future IPC handlers) can fire it without pulling in the hook.
After AI settings apply, `syncPayload.ts` walks every AI key it
touched (including the agentProviderMap / agentModelMap that the
reconcile step may have mutated even when the payload didn't ship
them) and emits a same-window nudge. `useAIState`'s existing local
listener routes unknown keys back through its storage handler, so
each affected piece of React state rehydrates from localStorage.
Adds a regression test in `syncPayload.test.ts` that asserts the
expected `netcatty:ai-state-changed` keys are dispatched for the
providers + per-agent map keys.
Refs codex local review on #1107.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
* Fix#954: unify Tooltip styling + replace native selects
Replace native HTML title= tooltips and native <select> dropdowns
with the existing Radix-based Tooltip / Select components so they
share the app's rounded styling, theme tokens and i18n pipeline.
Adds a global TooltipProvider in AppWithProviders so every
descendant Tooltip works without a per-file Provider wrapper.
Scope (driven by the issue #954 examples and "全部都处理" follow-up):
- TerminalLayer toolbar: Add Terminal / Split View / SFTP / Scripts
/ Theme / AI Chat / Move panel / Close panel.
- TopTabs middle bar: quick switcher, more tabs, AI assistant, theme
toggle, settings; window-control buttons (min/max/close), tray
close and hotkey reset/disable have their native title dropped per
the user's explicit opt-out ("可以不用Tooltip,直接全局禁用
原生title 属性").
- AI panels: AIChatSidePanel session history / new chat / delete,
ConversationExport, AgentSelector, ChatInput attach / expand /
permission, ModelSelector, ProviderCard, ai-elements/tool-call.
- SFTP: SftpSidePanel header, SftpBreadcrumb, SftpFileRow,
SftpPaneToolbar, SftpTabBar, SftpTransferQueue.
- Settings: SettingsPage close, SettingsAppearanceTab theme/accent
swatches, SettingsFileAssociationsTab edit/remove, SettingsSystemTab
crash-log paths and global hotkey reset.
- Host vault: HostDetailsPanel (clear / suggestions / show-password /
key path / browse key), GroupDetailsPanel, KnownHostsManager,
ConnectionLogsManager, KeychainManager, SyncStatusButton,
CloudSyncSettings, LogView, QuickSwitcher, ScriptsSidePanel,
Terminal status bar copy-host + broadcast/focus, ZmodemProgressIndicator.
- Terminal subcomponents: HostKeywordHighlightPopover, TerminalComposeBar,
TerminalConnectionDialog, TerminalSearchBar.
- Editor: TextEditorPane (subtitle, search, wrap, promote-to-tab).
- TrayPanel session rows and port-forwarding rows.
Native <select> migrated to custom Select component:
- SerialConnectModal (data bits, stop bits, parity, flow control)
- SerialHostDetailsPanel (same four fields)
- HostDetailsPanel backspace behavior
- GroupDetailsPanel backspace behavior
- SettingsTerminalTab local shell picker
- terminal/ThemeSidePanel font weight
Hardcoded English strings extracted to i18n. New keys for both
en and zh-CN: terminal.layer.*, topTabs.*, ai.chat.* (sessionHistory,
attach, collapse, expand, enableAgent), zmodem.*, settings.shortcuts.
resetToDefault. Inline help text on SnippetsManager package-name input
removed because the same hint is already shown in a visible <p> below
the input.
Existing per-file <TooltipProvider> wrappers (SnippetsManager,
ScriptsSidePanel, SelectHostPanel, RuleCard, HostDetailsPanel proxy
section) are left in place — they nest harmlessly under the global
provider and stay self-sufficient for component tests.
Tests:
- tsc clean for changed files (pre-existing repo-wide errors
unrelated to this PR).
- All 802 tests pass (3 skipped pre-existing).
- HostDetailsPanel.proxyProfile.test and TextEditorPane.test
updated to wrap with TooltipProvider, matching the runtime
context now needed by the migrated components.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
* Fix#954: wrap Settings + Tray windows with TooltipProvider
Settings and the tray panel mount as separate Electron windows with
their own React root in index.tsx, so they do not inherit the global
TooltipProvider added under AppWithProviders. After the unified
Tooltip migration, any settings tab that used a Tooltip (Appearance,
Application, FileAssociations, System, Shortcuts, Terminal, AI
ProviderCard, AI ModelSelector) — and TrayPanel — threw
"Tooltip must be used within TooltipProvider" and rendered nothing.
Wrap both branches with TooltipProvider at the same level as
ToastProvider in index.tsx.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
- Fixed 8% brightness causes compositers to have severe rendering issues. (Only effected on the Midnight color scheme) 10% seems to be okay.
- Reduced backdrop-blur as it's expensive CSS.
- Removed radial-gradient backgrounds (they don't show up)
* feat(ai-chat): fit-to-content popovers and keyboard nav for @/slash menus
- Shrink the @ host and /skill popovers to their content width
(auto width with min 220px, capped at the input width) instead of
always filling the full input width, which left large empty gutters
when the list was short.
- Add keyboard navigation: ArrowUp/ArrowDown cycle through items,
Enter commits the highlighted item, Escape closes the menu. Mouse
hover stays in sync with the active index so keyboard and pointer
agree on which row is current. Enter does not fall through to
submit while a menu is open.
- Expose aria-selected / aria-activedescendant for screen readers.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* style(ai-chat): tone down popover radius to match other menus
The @ and /skill popovers used rounded-[20px]/rounded-[16px] which
stood out against every other popover in this file (rounded-lg with
rounded-md items). Switch to the shared radii and drop shadow-2xl for
the standard shadow-lg so the surface feels consistent.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* style(ai-chat): tighten mention popover spacing
- Drop the redundant "Hosts" / "User Skills" header row — the @ or /
trigger already makes the popover's purpose obvious, and the header
added ~30px of vertical whitespace above a single-line list.
- Shrink wrapper and item padding (p-2.5/px-3 py-1.5 -> p-1/px-2 py-1)
and remove the mt-0.5 gap between title and subtitle.
- Hide the hostname subline when the label already contains the
hostname (common case: "Rainyun-114.66.26.174" as label and
"114.66.26.174" as hostname — no need to repeat).
- Lower minWidth 220 -> 200 so short lists can shrink further.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix(ai-chat): address Codex review on PR #726
- Reset active menu index on any change to the *set* of visible items,
not just its length. Watching only `.length` let Enter commit a
different item when the slash query changed to a same-sized match
set. Derive a stable identity key (sessionIds / skill ids) and use
that as the effect dep instead.
- Clamp the popover's minWidth to the measured panel width so narrow
layouts don't end up with minWidth > maxWidth, which CSS resolves
by honoring min and clips the menu off-screen.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- Extract fail-loud check to shared getCodexCustomConfigPreflightError so
the list-models handler (aiBridge.cjs:2149) enforces the same up-front
error as the stream handler. Previously a user whose config.toml
env_key was unexported would get the targeted message on chat send but
a generic "Missing env var" from model-list probes (once the probe was
rewired for Codex in a future change).
- Wire Settings "Refresh Status" to also invalidate the shell-env cache.
New invalidateShellEnvCache() helper in shellUtils; aiCodexGetIntegration
now accepts an optional { refreshShellEnv } flag; the button passes it
so a user who just exported OPENROUTER_API_KEY in their rc file can
click Refresh instead of having to restart netcatty.
- Declare authHash in CodexCustomProviderConfig (types.ts + global.d.ts)
so renderer TS actually sees the field instead of needing a cast.
- DRY the 360 magic number in ChatInput: extract
MODEL_PICKER_MAX_WIDTH, use it in both the className max-width and the
left-clamp math so the two can't drift.
- Move codexCustomConfigResolved useState declaration next to its
companion codexConfigModel, above the effect that invokes its setter,
and drop the duplicate declaration further down. Pure code-organization
cleanup but removes a use-before-declaration nit.
No functional changes beyond the fail-loud parity and the refresh-shell-env
path. ACP behavior when authMethodId is omitted still requires a
real-world OpenRouter config.toml validation, which the user is running.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Round of fixes driven by two parallel reviewers:
- i18n placeholder mismatch (P0). Locale strings used ${envKey} (literal
dollar-sign) but the replace call passed '{envKey}', so the warning
displayed a raw "${envKey}" instead of the real env var name. Align on
the codebase-standard {envKey} form.
- Fingerprint now folds the hash of the actual auth material (P1).
readCodexCustomProviderConfig computes a sha256 over the hardcoded
api_key or the resolved env_key value and returns authHash. The ACP
provider-reuse fingerprint includes it, so rotating the key in
~/.zshrc + restarting netcatty (which refreshes shellEnv) now
invalidates the cached provider instance instead of keeping the stale
key alive. Raw value never crosses the IPC boundary — we only send
the hex digest.
- Fail loud when config.toml's env_key isn't exported (P1). Previously
we'd sail into spawn and let codex-acp fail mid-request with a cryptic
"Missing environment variable". Now the stream handler rejects up
front with a targeted error naming the missing variable and pointing
at ~/.zshrc.
- TOML parser: basic-string escape tracking (P1). findUnquotedHash now
tracks an explicit `escaped` flag (and only honors escapes inside
double-quoted strings, since literal single-quoted strings don't), so
values like "C:\\path\\" close correctly instead of consuming the
trailing `#` as part of the string.
- TOML parser: strip UTF-8 BOM (P2). Windows editors frequently prepend
one and the first-key regex would silently fail to match, dropping
everything before the first section header.
- Picker correctness when config.toml lacks a `model` field (P1).
Instead of silently falling back to CODEX_MODEL_PRESETS (stock
OpenAI IDs the user's custom endpoint can't serve), show an empty
list so the picker disables. Track codexCustomConfigResolved so we
distinguish "still loading" from "not a custom-config session" and
only clear the preset list once the integration probe confirmed
connected_custom_config.
- Logout handler isConnected also considers connected_custom_config
(P2 consistency), matching get-integration.
- Model picker popover clamps its left position so max-w-[360px] can't
push it past the right edge of a narrow AI side panel (P2).
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Two issues the user flagged with the previous round:
1. Probing codex-acp for available models returned the stock ChatGPT
catalog (GPT 5.4, Codex 5.x, o3, o4-mini) regardless of the active
provider. For a user with a custom model_provider in
~/.codex/config.toml (OpenRouter + Qwen), those IDs are meaningless
on their endpoint. Roll back the managed-Codex probe hook and go
back to static CODEX_MODEL_PRESETS for the stock / ChatGPT path.
2. The fixed w-[300px] popover left empty space on the right whenever
the longest row was narrower than 300px.
Instead of the probe, teach readCodexCustomProviderConfig to also
return the top-level `model` from config.toml and expose it on the
integration response. In AIChatSidePanel, call aiCodexGetIntegration
when Codex is the active agent and, if customConfig.model is present,
override agentModelPresets with a single-entry list pinned to that
model. Otherwise fall back to the static presets as before — so
ChatGPT users see GPT 5.x / Codex 5.x etc. exactly like before, while
custom-config users see just the model their provider is actually
pinned to.
Popover switches from fixed width to `w-max min-w-[160px] max-w-[360px]`
so it hugs content (great for short single-model lists) while still
capping very long rows.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
The picker label was being derived by splitting selectedModelId on the
first '/'. That works for Codex's ChatGPT-preset format
("gpt-5.4/high" → model "gpt-5.4" + thinking level "high"), but breaks
for OpenRouter-style ids from config.toml ("qwen/qwen3.6-plus"):
selectedBaseModelId became "qwen", which doesn't match any preset, so
selectedPreset fell back to undefined and the chip displayed the
unrelated app-level modelName (e.g. "gemini-3-flash-preview") instead
of the actually selected Codex model.
Replace the naive split with a two-step lookup: first try a direct id
match; only if that fails, look for a preset whose declared
thinkingLevels make "${preset.id}/${level}" equal to selectedModelId,
and derive the thinking segment from that. Model ids that happen to
contain '/' now round-trip correctly through the picker.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
codex-acp's provider descriptions can be paragraphs ("Latest frontier
model with improvements across a wide range of capabilities..."), which
made each row of the picker feel bloated. The model id and (thinking
sub-menu's) thinking level already convey the relevant distinction —
drop the description render entirely. Keeps the dropdown tight regardless
of how verbose the upstream model catalog is.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Horizontal layout + truncate clipped too much of codex-acp's longer
descriptions ("Latest frontier model with improvements across a..." →
"Latest frontier model w..."). Reorganize each option as
checkmark | name-on-top, wrapped description below | chevron, so the
full description is readable across two lines without pushing the
popover width out. Fix popover to w-[300px] for a consistent column
width. Checkmark and chevron anchor to the first text line (self-start
with small top offset) so they stay visually aligned with the name
when the description wraps.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
With dynamic models now pulled from codex-acp, preset descriptions can be
arbitrarily long ("Latest frontier model with improvements across a..."
from OpenAI's public model list). The popover had whitespace-nowrap on
each option and no max-w on the container, so long descriptions pushed
the dropdown off-screen.
Cap the popover at max-w-[360px], add min-w-0 + truncate to the name
span so flex children can actually shrink, and cap the description span
at max-w-[160px] with truncate so it ellipses rather than expanding the
row. ChevronRight gets shrink-0 so it can't be pushed out of view.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix: preserve AI chat history across reconnects
* fix: retarget restored AI sessions on reconnect
* feat: format tool call results with proper line breaks
Extract stdout/stderr from structured results and unescape \n/\t
so command output displays with real line breaks like terminal output.
Supports both JSON object {stdout,stderr} and executor text formats.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix: restrict unescape to stdout/stderr fields only
Plain strings may contain legitimate backslash sequences (file paths,
regex patterns) that should not be converted. Only apply unescape to
stdout/stderr fields extracted from command execution results.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix: address review findings for AI chat reconnect
1. Add explicit activeTerminalTargetIds guard in shouldRetargetActiveSession
to prevent retargeting sessions owned by other terminals, making the
invariant locally verifiable.
2. Only preserve orphaned terminal sessions with hostIds — workspace,
local, and serial sessions generate fresh IDs and would be permanently
unreachable, wasting MAX_STORED_SESSIONS quota.
3. Clear stale streaming state when restoring a session whose ACP handle
was already cleaned up (e.g., reconnect during mid-response), so the
user can send new messages.
4. Restore overflow-hidden on user message bubbles to prevent content
bleeding past rounded border corners.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix: address round 2 review findings
1. Fix streaming state clear: only clear for sessions whose targetId
doesn't match current scope (restored from different terminal),
not for built-in Catty chats that never set externalSessionId.
2. Exclude local/serial sessions from preservation: their synthetic
hostIds (local-*/serial-*) change on every open and can never be
matched back.
3. Preserve non-zero exitCode in formatted tool results so failed
commands show a visible failure signal.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix: only clear streaming state during retarget, not for all restored sessions
The previous condition (targetId !== scopeTargetId) also fired for
built-in Catty sessions during normal operation, killing active streams.
Now streaming is only cleared when shouldRetargetActiveSession is true,
meaning the session came from a disconnected terminal where any
in-flight response is guaranteed to be dead.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix: address round 3 review findings
1. Clear externalSessionId during retarget to prevent stale ACP handle
from surviving if retarget runs before orphan cleanup.
2. Only retarget in visible AI panels — hidden/background panels should
not race to claim orphaned sessions.
3. Remove unescapeTerminalOutput — data flow trace confirms real newline
characters arrive at the component. The unescape was corrupting
legitimate backslash sequences in paths and patterns.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix: only ACP-cleanup deleted sessions, not preserved ones
Preserved sessions may be reused on reconnect. Running aiAcpCleanup
on them asynchronously could race with a newly started ACP conversation
on the same session ID, tearing down the fresh provider.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix: abort in-flight streams during retarget and restore ACP cleanup
1. Abort the active request's AbortController when retargeting a session
with stale streaming state. Prevents late chunks from the old run
appending into the restored chat.
2. Restore ACP cleanup for all orphaned sessions (not just deleted ones).
Preserved sessions get a new externalSessionId on next use, so
cleaning the old one prevents subprocess leaks without affecting
future conversations.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix: guard hidden panels from session ownership and skip null map entries
1. Only assign restored sessions in visible panels — hidden panels
should not race to own sessions via setActiveSessionId, preventing
MCP/tool calls from being bound to the wrong terminal.
2. Skip null entries in activeSessionIdMap when building
activeTerminalTargetIds — deleted chats should not block same-host
history matching on other terminals.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix: guard MCP sync behind visibility and cancel exec/approvals on retarget
1. Only sync MCP session metadata from visible panels to prevent
hidden panels from overwriting the scope mapping.
2. Cancel pending approvals and in-flight exec (Catty + ACP) during
retarget, matching handleStop behavior. Prevents stale tool results
and approval prompts from reappearing after session retarget.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix: restore MCP sync for hidden panels
MCP scope is keyed by chatSessionId so hidden panels don't overwrite
visible panels' mappings. The isVisible guard was breaking background
chats that need updated terminal session metadata after reconnects
or workspace changes.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* chore: remove unused deletedIds variable
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: bincxz <16399091+binaricat@users.noreply.github.com>
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* feat: support multimodal attachments (images, PDFs, files) in AI chat
Previously uploaded images were displayed in the UI but never sent to
the AI model, and non-image files (PDF, text) were silently rejected.
- Rename useImageUpload → useFileUpload; accept image/*, PDF, and text/*
- Rename ChatMessageImage → ChatMessageAttachment with filePath support
- Build multimodal SDK messages (ImagePart/FilePart) for Catty Agent
- Fix ACP agent path: images inline, non-image files via local path hint
so ACP agents (Claude Code, etc.) read them with native file access
- Use Electron webUtils.getPathForFile() for reliable file path capture
- Compact user message bubble padding
Closes#294 (AI file upload issues)
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix: show real tool names in AI chat instead of ACP wrapper names
- Unwrap ACP dynamic tool calls in serializeStreamChunk to extract
real tool name, args, and toolCallId from chunk.input
- Simplify MCP tool name prefixes (mcp__server__tool → tool)
- Pass toolCallId from ACP tool-call events to match tool results
- Prevent onToolResult from overwriting correct names with wrapper name
- Build toolCallNames map in ChatMessageList for tool result display
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix: backward-compatible fallback for legacy `images` field in chat messages
Persisted sessions may still have `images` instead of `attachments`.
Add `?? m.images` fallback in SDK message builder and renderer so
historical image attachments are not silently dropped after upgrade.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix: broaden file type support and handle pasted files without path
- Accept all file types except video/audio (instead of allowlist)
so .json, .yaml, .sh, etc. are not silently rejected
- For ACP agents, save pasted/virtual files (no filePath) to temp
directory so the agent can still read them
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix: use managed temp dir for pasted ACP attachments
Use tempDirBridge.getTempFilePath() instead of manual os.tmpdir() path
so pasted file attachments are tracked by the app's cleanup system.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- Security: API keys no longer transit IPC as plaintext; renderer sends
providerId and main process decrypts via safeStorage. Add ReDoS
protection for user-supplied blocklist regex. Sanitize error messages
to strip file paths and sensitive URLs before displaying in chat.
- Bug fixes: approval timeout now notifies user in chat instead of
silently aborting. statusText cleared consistently across all 7 code
paths. useAIState persistence race condition fixed with mountedRef
guard, storage sync validated with type checks.
- Accessibility: InlineApprovalCard uses role="alertdialog" with focus
on approve button. ChatInput menus have proper ARIA roles, labels,
and aria-expanded. ThinkingBlock toggle has aria-expanded/controls.
Model selector submenu supports keyboard navigation.
- UX: switching agents preserves old session and creates new one.
Approval card buttons disabled immediately after click.
- Performance: text-delta streaming batched via requestAnimationFrame.
- Refactor: extract useAIChatStreaming, useToolApproval, and
useConversationExport hooks from AIChatSidePanel (1514 → 751 lines).
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Security:
- Apply checkCommandSafety() in aiExec IPC handler to enforce command blocklist
- Add critical path validation in handleSftpRemove (normalize + blocklist)
- Change rm -rf to rm -r so permission errors surface
Stream lifecycle:
- Abort all active streams on component unmount (abortControllersRef cleanup)
- Wrap stream reader in try/finally with releaseLock() to prevent leaks
- Use refs for inputValue/images in handleSend to stabilize callback identity
State persistence:
- Clear debounce timer before synchronous persist in destructive operations
(clearSessionMessages, deleteSession, deleteSessionsByTarget)
- Add 10MB max buffer guard in ACP client and MCP server NDJSON parsing
i18n:
- Replace hardcoded English strings with t() calls in InlineApprovalCard,
PermissionDialog, ConversationExport, ThinkingBlock, AgentSelector
- Add 23 new i18n keys to en.ts and zh-CN.ts
Misc:
- Remove debug console.log statements in mcpServerBridge
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Show a clickable chip next to the model selector in ChatInput footer
that lets users quickly toggle between Observer/Confirm/Auto permission
modes. Only visible when Catty Agent is selected (ACP agents handle
their own tool approval flow).
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Safety enforcement:
- Command Timeout: use user setting in MCP Server (execViaPty/execViaChannel)
and Catty Agent path (aiBridge execViaPtyForCatty/execViaChannel fallback)
- Max Iterations: read user setting for Catty (stepCountIs) and ACP (via IPC)
- Permission Mode: observer mode hard-blocks write ops in MCP Server dispatch;
Catty SDK tools wired to checkToolPermission(); all synced via IPC on change
Command Blocklist UI:
- Add editable regex pattern list in Settings AI Safety section
- Reset to defaults button, add/remove patterns
- IPC sync to MCP Server on change and on mount
Provider UX:
- ModelSelector rewritten as combobox: type-to-filter with suggestions dropdown
- All providers use unified ModelSelector (modelsEndpoint optional)
- API key passed as auth header for model fetching (Bearer / x-api-key)
- Skip model fetch when no API key (except Ollama)
- Provider toggle is now mutually exclusive (activating one disables others)
- New providers default to disabled (switch off)
- Custom provider supports editable display name
- No-provider error: friendly message shown in chat
i18n:
- Full i18n coverage for Settings AI tab (~55 keys)
- Full i18n for AI chat panel (placeholder, empty state, time, sessions)
- en.ts and zh-CN.ts locale files updated
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
AI Settings:
- Remove External Agents section, add dedicated Codex/Claude Code sections
with auto-detection and manual path override
- Add OpenRouter model list auto-fetch with searchable dropdown
- Remove standalone Default Model section, integrate into provider cards
- Provider toggle now acts as mutual-exclusive active selector
- Add cross-window state sync via storage events
- Add ErrorBoundary around AI tab for graceful error handling
Catty Agent fixes:
- Fix streaming: use .chat() instead of default Responses API, use
getReader() pattern instead of for-await, handle text-delta/text chunk
types correctly per SDK v6
- Fix API key: decrypt encrypted key before passing to SDK
- Fix terminal_execute: use PTY stream (visible in terminal) with MCP
markers instead of invisible SSH exec channel
- Fix multi-turn: only pass user/assistant text history, skip tool
messages to avoid SDK schema validation errors
- Fix tool result display: create new assistant message after tool
results so follow-up text renders correctly
Other:
- Add netcatty:ai:resolve-cli IPC handler for CLI path validation
- Remove Gemini from agent discovery (only Codex + Claude Code)
- Fix lint errors across ChatInput, TerminalLayer, SettingsAITab
- Strip MCP markers from tool execution stdout
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- Add image paste/drop support with base64 encoding and chat display
- Add + button popover (Files, Image, Mention Host) with @ auto-complete
- Add model selector with Codex thinking level sub-menus (GPT 5.4, Codex 5.x, o3/o4-mini)
- Switch Claude Code to ACP protocol; remove CLAUDE_CONFIG_DIR isolation
- Filter PTY exec markers in preload data pipe (precise regex, preserves command echo)
- Increase PTY exec timeout to 5min with Ctrl+C cancellation on abort
- Fix tool call loading spinner (only animate during active streaming)
- Reset active session on terminal/workspace switch
- Add AI sparkle button in top bar to toggle AI panel
- Display user-attached images in chat message list
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- Create netcatty-remote-hosts MCP server (TCP bridge + stdio child process)
exposing terminal_execute, sftp_*, get_environment tools to ACP agents
- Execute commands via PTY stream with self-erasing markers for terminal
visibility; disable pagers automatically; 60s timeout fallback
- Inject MCP server into both Codex (ACP) and Claude Code (ACP) sessions
- Add model selector popover with hardcoded presets for Claude (Opus/Sonnet/
Haiku) and Codex (GPT 5.4, Codex 5.3/5.2/5.1, o3/o4-mini) with thinking
level sub-menus
- Fix multi-step tool call message flow (mutable flag instead of stale closure)
- Remove CLAUDE_CONFIG_DIR isolation to fix Claude auth; switch Claude to ACP
- Add startup cleanup for orphaned AI sessions
- Guard collectSessionIds against undefined workspace tree
- Remove permission mode chip from chat input
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Integrate Claude Agent SDK for direct streaming chat, add Codex login/logout
flow with OAuth support in settings, improve AI chat panel UI and agent
discovery, and update build config for new dependencies.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>