openclaw/docs/platforms/mac/webchat.md at da69daeb72cf47fb76afe55e5216319fd4245933

mirror of https://github.com/openclaw/openclaw.git synced 2026-07-22 19:21:17 +00:00

Files

Peter Steinberger bf754b8a34 feat: align native macOS chat with the web chat capability set (#109712 )

* feat(macos): wire sessions manager, assistant trace, talk, voice notes, and Listen into native chat

* feat(clients): add input history, reply quoting, and per-session drafts to native chat composer

* fix(clients): capture TTS server lease before gateway check and dedupe restored input-history entries

* docs(clients): document reply-chip consume parity with web accepted-send semantics

* fix(clients): retire sent-text recall stash and keep recall markers stable under prepends

* test(clients): wait for outbox restore before asserting live-send paths

* docs(clients): note cross-session outbox flush scheduling contract

* chore(clients): split oversized chat files to satisfy SwiftLint length caps

* chore(i18n): sync native app i18n inventory for new chat strings

* chore: retrigger ci

* chore(i18n): translate new chat strings and regenerate Apple catalogs

* chore(i18n): regenerate Android localization resources for new chat strings

* test(clients): await lifecycle cleanup before modeling replacement gateway traffic

* chore(clients): split invoke timeout helper to satisfy SwiftLint length cap

* chore(clients): drop unused OpenClawProtocol import from invoke timeout split

2026-07-17 13:44:48 +01:00

4.3 KiB

Raw Blame History

summary, read_when, title

summary

read_when

title

How the mac app embeds the gateway WebChat and how to debug it

Debugging mac WebChat view or loopback port

WebChat (macOS)

The macOS menu bar app embeds the WebChat UI as a native SwiftUI view. It connects to the Gateway and defaults to the primary session for the selected agent (main, or global when session.scope is global).

The full chat window is a native split view:

Sessions sidebar: searchable session list with pinned and recent sections, unread indicators, and context menus for pin/unpin, copy session key, and delete. A toolbar button (or Cmd-N) creates a real new session via sessions.create.
Window toolbar: context-usage ring (tokens and session cost, with a compact action), thinking-level picker, model picker, and a session actions menu. Sessions… (Shift-Cmd-S) opens the Active/Archived manager for gateway search, rename, pin, archive, and restore. The same menu can show or hide assistant reasoning and tool activity; this is on by default and remembered across launches.
Transcript and composer: assistant messages render as plain text with an avatar, user messages as accent bubbles. Empty chats offer desktop starter prompts. Typing / opens slash-command autocomplete backed by commands.list, with arrow/Tab/Return/Escape keyboard navigation. Right-click a message to copy it, or use Listen for gateway TTS with a local speech fallback.
Voice controls: the composer can start or stop the existing macOS Talk Mode without replacing its menu-bar overlay. A separate microphone action records a voice note when Talk Mode does not own audio capture.

The anchored compact chat panel from the menu bar keeps the compact single-column layout with inline pickers, starter prompts, Talk Mode, voice notes, and Listen. Assistant reasoning and tool activity remain hidden in this compact surface.

Quick Chat bar

Press Option-Space (⌥Space) or choose Quick Chat from the menu bar menu to open a floating composer for the main session. Change the global shortcut with the recorder in Settings → General → Quick Chat shortcut.

Quick Chat shows the main session's agent, sends directly to the main session, and leaves replies in the full chat window. Press Return to send, Command-Return to send and open full chat, Shift-Return for a newline, or Escape to dismiss. Clicking outside the bar also dismisses it. When relevant macOS permissions are missing, an attached strip offers Grant and Not now actions.

Local mode: connects directly to the local Gateway WebSocket.
Remote mode: forwards the Gateway control port over SSH and uses that tunnel as the data plane.

Launch and debugging

Manual: Lobster menu -> "Open Chat".
Auto-open for testing:
```
dist/OpenClaw.app/Contents/MacOS/OpenClaw --chat
```
(--webchat is accepted as a legacy alias.)
Logs: ./scripts/clawlog.sh (subsystem ai.openclaw, category WebChatSwiftUI).

How it is wired

Data plane: Gateway WS methods chat.history, chat.send, chat.abort, chat.inject, and events chat, agent, presence, tick, health.
chat.history returns a display-normalized transcript: inline directive tags are stripped from visible text, plain-text tool-call XML payloads (<tool_call>, <function_call>, <tool_calls>, <function_calls>, including truncated blocks) and leaked model control tokens are stripped, pure silent-token assistant rows such as exact NO_REPLY/no_reply are omitted, and oversized rows can be replaced with a truncated placeholder.
Session: defaults to the primary session as above; the UI can switch between sessions.
Onboarding uses a dedicated session to keep first-run setup separate.
Offline cache: the app keeps a small read-only cache of recent chat sessions and transcripts per gateway (~/Library/Application Support/OpenClaw/chat-cache.sqlite): cold opens paint the last known transcript immediately and refresh once the Gateway responds, and recent chats stay browsable while disconnected (sending stays disabled until the connection is back).

Security surface

Remote mode forwards only the Gateway WebSocket control port over SSH.

Known limitations

The UI is optimized for chat sessions, not a full browser sandbox.

4.3 KiB Raw Blame History