comate-cli 0.10.22__tar.gz → 0.11.2__tar.gz

{comate_cli-0.10.22 → comate_cli-0.11.2}/CHANGELOG.md

@@ -18,6 +18,9 @@
 
 ### Added
 
+- 新增 `/workflows` 独立全屏面板：跑次列表 → 跑次详情（阶段标题 + agent 行）→ agent 详情
+  （prompt / 结果 / 错误 pager）三层导航，每秒轮询 run JSON live 刷新；`x` 对本会话发起的
+  跑次两段式停止，非本会话 / 崩溃遗留的 running 行标记为「非本会话」且不可停；忙时可打开。
 - `comate mcp serve`：progress 通知携带工具参数与 subagent 内部工具调用（此前只有裸工具名）。
 - TUI jobs 面板：后台 agent 的当前工具调用实时显示在 running 行内（label 后 5 空格、暗色、原地替换）。
 - TUI 折叠容器：动态展示下游 MCP server 推送的 progress 通知（嵌套暗色行单行原地替换，调用结束即消失）。

{comate_cli-0.10.22 → comate_cli-0.11.2}/CONTEXT.md

@@ -53,8 +53,8 @@ The first non-empty user-authored input accepted as durable session work. Reject
 _Avoid_: first keypress, submitted text (ambiguous)
 
 **Durable Title Completion**:
-A successfully persisted title for a Session Report. Resuming the same session reuses this fact instead of generating another title.
-_Avoid_: generated-title flag, controller completion
+The fact that a Session Title has been persisted to the session's own metadata. Resuming the same session reuses this fact instead of generating another title; a Session Report only mirrors it.
+_Avoid_: generated-title flag, controller completion, report title
 
 **Usage Ledger**:
 The session's monotonic accumulated record of actual LLM usage. It is the source of truth for the token usage projected into a Session Report.
@@ -134,6 +134,20 @@ _Avoid_: discard input (the pre-existing lossy clear), undo (there is no in-plac
 A short-lived hint occupying the status bar's left column, dismissed by timeout or the next real keypress, whichever comes first. It is interaction feedback, not history: it never enters scrollback.
 _Avoid_: toast, notification, system message (that is the scrollback form)
 
+## Terminal Tab Language
+
+**Session Title**:
+A short (3–7 word) title generated once from the session's earliest user input, owned by the session itself and shared by the resume picker, the Tab Title, and the Session Report. Once a title exists — generated or manually named — automatic generation never replaces it.
+_Avoid_: summary, session summary, report title, auto title (that names the mechanism, not the fact)
+
+**Tab Title**:
+The single line of text the terminal shows in its window or tab for a running interactive CLI: an Attention State glyph followed by the Session Title, or by the product name while no Session Title exists. It reaches the terminal through two channels carrying the same text — the terminal title escape sequence and the foreground process name — because some terminals (VS Code by default) name tabs after the process rather than the sequence. It is never written in non-interactive (print / rpc) mode.
+_Avoid_: window title, OSC title, process title (each names one channel, not the fact)
+
+**Attention State**:
+What the session currently needs from the user, as one of three values: idle (nothing running), busy (a turn or long command is running), needs-you (execution is paused on a tool approval or a question to the user). It is the only input that selects the Tab Title glyph.
+_Avoid_: spinner state, running flag, blocked
+
 ## Relationships
 
 - A **Core Slash Command** is always registered with `source="builtin"` on the Execution Axis. Its name belongs to the **Reserved Name Set** and cannot be overridden by any markdown-template command.
@@ -165,6 +179,12 @@ _Avoid_: toast, notification, system message (that is the scrollback form)
 - The **Update Effectiveness Boundary** holds for both Plugin Auto Update and manual plugin updates: neither may mutate the running session's loaded resources.
 - An **Auto Update Opt-out** affects only Plugin Auto Update; it does not disable manual per-plugin updates.
 
+- A **Session Title** is a fact of the session, not of the **Session Report**: it exists whether or not a report backend is configured, and the report only mirrors it. **Durable Title Completion** is therefore judged from the session's own metadata.
+- A **Session Title** is generated from the session's earliest **Committed Title Input**, never from the latest one. A resumed session without a title is titled from its original first input, so the title does not change to whatever was typed after resume.
+- The **Tab Title** shows the **Session Title** the moment it becomes durable, falls back to the product name after `/clear` starts a new session, and shows the existing title immediately on resume.
+- **Attention State** follows the TUI's single busy switch plus the two pause points (tool approval, question to the user). Background job completion and the exit confirmation dialog do not change it.
+- **Attention State** never animates: each state maps to exactly one glyph, so the terminal is written only on state changes and title changes.
+
 ## Flagged ambiguities
 
 - None currently. The original "builtin" collision between Execution Axis and Source Axis has been resolved by reserving "builtin" for Core Slash Commands and introducing "default" for the new shipped markdown resources.

{comate_cli-0.10.22 → comate_cli-0.11.2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: comate-cli
-Version: 0.10.22
+Version: 0.11.2
 Summary: Comate terminal CLI built on comate-agent-sdk
 Project-URL: Homepage, https://github.com/AndyLee1024/agent-sdk
 Project-URL: Repository, https://github.com/AndyLee1024/agent-sdk
@@ -15,7 +15,7 @@ Classifier: Programming Language :: Python :: 3.12
 Classifier: Programming Language :: Python :: 3.13
 Requires-Python: <3.14,>=3.11
 Requires-Dist: charset-normalizer==3.4.7
-Requires-Dist: comate-agent-sdk<0.13,>=0.12.4
+Requires-Dist: comate-agent-sdk<0.13,>=0.12.5
 Requires-Dist: concurrent-log-handler>=0.9.25
 Requires-Dist: curl-cffi==0.13.0
 Requires-Dist: httpx<0.29.0,>=0.28.1

{comate_cli-0.10.22 → comate_cli-0.11.2}/comate_cli/main.py

@@ -552,7 +552,10 @@ def _build_wellformed_input():
 def main(argv: list[str] | None = None, *, profile=None) -> None:
     from comate_cli.distribution import DEFAULT_PROFILE
     from comate_cli.terminal_title import (
+        AttentionState,
         clear_terminal_title,
+        format_tab_title,
+        process_title_text,
         set_process_title,
         set_terminal_title,
     )
@@ -618,7 +621,13 @@ def main(argv: list[str] | None = None, *, profile=None) -> None:
     if interactive:
         from prompt_toolkit.application import create_app_session
 
-        set_terminal_title(command_name)
+        # 启动期先写「空闲 + 产品名」占位；TUI 起来后由它接管 Tab Title。
+        # OSC 与进程名两条通道同一串文字（VS Code 默认按进程名命名 tab）。
+        startup_tab_title = format_tab_title(
+            AttentionState.IDLE, None, product_name=command_name
+        )
+        set_process_title(process_title_text(startup_tab_title))
+        set_terminal_title(startup_tab_title)
         input_session = create_app_session(input=_build_wellformed_input())
     else:
         input_session = nullcontext()

{comate_cli-0.10.22 → comate_cli-0.11.2}/comate_cli/terminal_agent/app.py

@@ -1314,16 +1314,13 @@ async def run(
 
     # TerminalAgentTUI 内部构造 prompt_toolkit Application；保持在 rich Live
     # spinner 窗口之外，避免任何 output 探测命中 FileProxy。
-    auto_title_controller = None
-    if report_controller is not None:
-        from comate_cli.terminal_agent.backend.session_title import (
-            AutoTitleController,
-        )
+    # Session Title 是会话事实：不依赖后端服务，report 只做镜像（可为 None）。
+    from comate_cli.terminal_agent.backend.session_title import AutoTitleController
 
-        auto_title_controller = AutoTitleController(
-            session=session,
-            report=report_controller,
-        )
+    auto_title_controller = AutoTitleController(
+        session=session,
+        report=report_controller,
+    )
     tui: Any
 
     async def _exit_epilogue() -> None:

{comate_cli-0.10.22 → comate_cli-0.11.2}/comate_cli/terminal_agent/backend/session_report.py

@@ -568,6 +568,14 @@ class SessionReportController:
             )
         except asyncio.CancelledError:
             raise
+        except FileNotFoundError:
+            # total_usage.json 只在某轮成功结束后才写出；首轮就失败的会话没有这个文件，
+            # 这是正常状态，不是故障，按空账本处理即可。
+            logger.debug(
+                "session report total usage not written yet: session_id=%s",
+                self.state.session_id,
+            )
+            return TotalUsageSnapshot.new(self.state.session_id)
         except Exception:
             logger.warning(
                 "session report total usage load failed: session_id=%s",

comate_cli-0.11.2/comate_cli/terminal_agent/backend/session_title.py

@@ -0,0 +1,176 @@
+"""会话标题（Session Title）自动生成控制器。
+
+Session Title 是**会话的事实**，不是上传报告的事实：无论有没有后端服务，
+首轮结束后都生成一次，写进 SDK 会话元数据（resume 选择器、Tab Title 共用
+这一份）；Session Report 只做镜像。
+
+不变量：
+- 已有标题（自动生成或手动改名）绝不覆盖——写前一刻再查一次（write-if-empty），
+  与 openworker ``server/titles.py`` 语义一致。
+- 取材固定为会话最早那条用户输入：resume 进来的老会话没标题时，用会话
+  首条消息补生成，而不是 resume 后新发的那句。
+"""
+
+from __future__ import annotations
+
+import asyncio
+import logging
+from pathlib import Path
+from typing import TYPE_CHECKING, Callable
+
+from comate_agent_sdk.agent.session_store import (
+    load_session_metadata,
+    rename_session,
+)
+from comate_agent_sdk.agent.session_title import (
+    SessionTitleError,
+    first_user_message_text,
+    generate_session_title,
+    normalize_session_title,
+)
+
+if TYPE_CHECKING:
+    from comate_agent_sdk.agent.chat_session import ChatSession
+    from comate_cli.terminal_agent.backend.session_report import (
+        SessionReportController,
+    )
+
+logger = logging.getLogger(__name__)
+
+_USER_TURN_SOURCES = frozenset({"direct_user", "queued_boundary"})
+
+#: 标题成为 durable 事实后的通知回调 ``(title) -> None``（TUI 用它刷新 Tab Title）。
+OnTitleFn = Callable[[str], None]
+
+__all__ = [
+    "AutoTitleController",
+    "SessionTitleError",
+    "current_session_title",
+    "generate_session_title",
+    "normalize_session_title",
+]
+
+
+def current_session_title(session: "ChatSession") -> str | None:
+    """会话元数据里当前的标题；没有元数据（非持久化会话 / 假会话）视为无标题。"""
+    storage_root = getattr(session, "_storage_root", None)
+    if storage_root is None:
+        return None
+    metadata = load_session_metadata(Path(storage_root))
+    if metadata is None:
+        return None
+    title = metadata.get("title")
+    return title if isinstance(title, str) and title.strip() else None
+
+
+class AutoTitleController:
+    def __init__(
+        self,
+        *,
+        session: "ChatSession",
+        report: "SessionReportController | None" = None,
+        on_title: OnTitleFn | None = None,
+    ) -> None:
+        self._session = session
+        self._report = report
+        self.on_title = on_title
+        # resume 的老会话：取材钉死在会话最早那条用户输入。
+        self.first_user_input: str | None = first_user_message_text(session)
+        self.current_turn_is_user_authored = False
+        self.attempts = 0
+        self._task: asyncio.Task[None] | None = None
+        self._closed = False
+        self.completed = current_session_title(session) is not None
+
+    @property
+    def task(self) -> asyncio.Task[None] | None:
+        return self._task
+
+    def record_user_input(self, text: str) -> None:
+        if self.first_user_input is not None:
+            return
+        normalized = str(text).strip()
+        if normalized:
+            self.first_user_input = normalized
+
+    def record_turn_started(self, source: str) -> None:
+        self.current_turn_is_user_authored = str(source) in _USER_TURN_SOURCES
+
+    def on_turn_finished(self, *, had_assistant_output: bool) -> None:
+        is_user_turn = self.current_turn_is_user_authored
+        self.current_turn_is_user_authored = False
+        if (
+            not is_user_turn
+            or self._closed
+            or not had_assistant_output
+            or self.first_user_input is None
+            or self.completed
+            or self.attempts >= 2
+            or (self._task is not None and not self._task.done())
+        ):
+            return
+        self.attempts += 1
+        self._task = asyncio.create_task(
+            self._run_attempt(),
+            name="terminal-session-auto-title",
+        )
+
+    def cancel(self) -> None:
+        self._closed = True
+        task = self._task
+        if task is not None and not task.done():
+            task.cancel()
+
+    async def _run_attempt(self) -> None:
+        title: str | None = None
+        try:
+            title = await generate_session_title(
+                self._session,
+                user_text=self.first_user_input or "",
+            )
+        except asyncio.CancelledError:
+            raise
+        except Exception:
+            logger.debug("session title generation failed", exc_info=True)
+        finally:
+            try:
+                self._session.flush_usage_ledger()
+            except Exception:
+                logger.debug("session title usage flush failed", exc_info=True)
+
+        if title is None:
+            return
+        self.completed = True
+
+        # write-if-empty：生成期间用户可能已手动改名，写前一刻再查。
+        existing = current_session_title(self._session)
+        if existing is not None:
+            title = existing
+        else:
+            self._persist_title(title)
+
+        if self._report is not None:
+            try:
+                await self._report.refresh_title(title)
+            except asyncio.CancelledError:
+                raise
+            except Exception:
+                logger.debug("session title report refresh failed", exc_info=True)
+
+        if self.on_title is not None:
+            try:
+                self.on_title(title)
+            except Exception:
+                logger.debug("session title on_title callback failed", exc_info=True)
+
+    def _persist_title(self, title: str) -> None:
+        session_id = getattr(self._session, "session_id", None)
+        cwd = getattr(self._session, "_cwd", None)
+        if not session_id or cwd is None:
+            logger.debug("session title not persisted: session has no storage identity")
+            return
+        try:
+            rename_session(session_id, cwd=cwd, title=title)
+        except Exception:
+            # 非持久化会话没有 session.json；标题仍在内存里供 Tab Title 使用。
+            logger.debug("session title metadata write failed", exc_info=True)

{comate_cli-0.10.22 → comate_cli-0.11.2}/comate_cli/terminal_agent/figures.py

@@ -76,6 +76,7 @@ LOCK_ICON = "🔒"             # 🔒
 # Blockquote / 装饰横线
 BLOCKQUOTE_BAR = "▎"         # ▎
 HEAVY_HORIZONTAL = "━"       # ━
+LIGHT_HORIZONTAL = "\u2500"  # ─
 
 # Bridge 状态指示
 BRIDGE_SPINNER_FRAMES: tuple[str, ...] = (

{comate_cli-0.10.22 → comate_cli-0.11.2}/comate_cli/terminal_agent/history_printer.py

@@ -507,12 +507,12 @@ def render_history_group(
         if entry.entry_type == "tool_fold":
             # 折叠组摘要：参考 thinking 的视觉处理（2 空格缩进 + dim 文案），
             # 不带前缀圆点，行末追加 "(ctrl+o to expand)" 提示 Ctrl+O 展开。
-            # 组内含失败时尊重 event_renderer 已算出的 severity，失败不再隐形；
-            # 展开提示后缀在任何状态下保持 faint。
+            # 失败可见性交由实时面板的红色 dot 与 viewer 的红色 bullet，
+            # scrollback 摘要行保持中性灰，避免高频软错误把整段工具调用上下文染红。
             content = str(entry.text).strip()
             content_lines = [line for line in content.splitlines() if line.strip()] or [content]
             suffix = " (ctrl+o to expand)" if entry.tool_call_ids else ""
-            text_style = theme.ERROR if entry.severity == "error" else theme.TEXT_FAINT
+            text_style = theme.TEXT_FAINT
 
             line_text = Text()
             line_text.append("  ", style=text_style)

{comate_cli-0.10.22 → comate_cli-0.11.2}/comate_cli/terminal_agent/i18n/_en.py

@@ -120,6 +120,20 @@ EN: dict[str, str] = {
         "or Ctrl+C again to exit."
     ),
     "mode_switch_failed": "Failed to switch permission mode: {error}",
+    "output_style_cancelled": "Output style change cancelled.",
+    "output_style_invalid_argument": "Unknown output style {value}. Use one of: default, concise, explanatory, learning, auto, or a .md file path.",
+    "output_style_menu_context_current": "Current style: {name} (source: {source})",
+    "output_style_menu_context_soul_detected": "SOUL.md detected: {path}",
+    "output_style_menu_context_soul_overridden": "SOUL.md detected ({path}), but overridden by an explicit selection.",
+    "output_style_menu_title": "Select output style",
+    "output_style_option_concise_desc": "Concise: only what changes your next step",
+    "output_style_option_default_desc": "Default style (outcome-first, concise)",
+    "output_style_option_explanatory_desc": "Explanatory: explains design trade-offs while working",
+    "output_style_option_learning_desc": "Learning: leaves small pieces of code for you to write",
+    "output_style_option_soul_desc": "Follow SOUL.md ({name})",
+    "output_style_switch_failed": "Failed to switch output style: {error}",
+    "output_style_switched": "Output style: {previous} → {new}",
+    "output_style_unchanged": "Output style unchanged ({name}).",
     "model_menu_effort_auto": "auto effort · /effort to change",
     "model_menu_effort_hint": "{glyph} {level} effort · /effort to change",
     "model_menu_effort_unsupported": "Effort not supported for {model}",
@@ -136,6 +150,12 @@ EN: dict[str, str] = {
     "plugin_auto_update_retry_ok": "Updated: {plugin_key}",
     "plugin_auto_update_toggled_off": "Auto update disabled: {name}",
     "plugin_auto_update_toggled_on": "Auto update enabled: {name}",
+    "preflight_endpoint_protocol_mismatch": (
+        "Endpoint protocol mismatch for {level}: provider '{provider}' speaks the {provider_protocol} protocol, "
+        "but base_url {base_url} looks like an endpoint for the {endpoint_protocol} protocol. "
+        "Check that llm_levels and llm_levels_base_url in settings.json match."
+    ),
+    "ps_agent_metrics": '{tokens} · {tool_calls} tools',
     "ps_unavailable_local_ui": (
         "/ps is not available while another local UI is open."
     ),
@@ -201,6 +221,60 @@ EN: dict[str, str] = {
     "transcript_empty": "No transcript entries yet. Send a message first.",
     "transcript_viewer_failed": "Transcript viewer failed: {error}",
     "webfetch_fetched": "Fetched {url} ∙ {chars} chars",
+    "workflow_approval_agent_limit": "agent calls: up to {max_calls}, {concurrency} in parallel; inner agents are not approved one by one",
+    "workflow_approval_no_phases": "phases: none declared in the script",
+    "workflow_approval_phases": "phases: {phases}",
+    "workflow_approval_resume": "resumes run {run_id}: finished agent calls are replayed",
+    "workflow_approval_script_lines": "script: {line_count} lines, saved to disk once started",
+    "workflow_approval_title": "Workflow run",
+    "workflow_summary_resume": "resume {run_id}",
+    "workflow_summary_script_lines": "{lines} lines of script",
+    "workflows_agent_metrics": '{model} · {elapsed} · {tokens} · {tool_calls}',
+    "workflows_agent_running": 'running · {elapsed} elapsed',
+    "workflows_agent_status_cached": 'cached',
+    "workflows_agent_status_done": 'done',
+    "workflows_agent_status_failed": 'failed',
+    "workflows_agent_status_running": 'running',
+    "workflows_agent_status_skipped": 'skipped',
+    "workflows_agent_title": 'Agent · {label}',
+    "workflows_empty": 'No workflow runs in this session yet.',
+    "workflows_footer_agent": '↑/↓ scroll · PgUp/PgDn page · ←/Esc back · q close',
+    "workflows_footer_list": '↑/↓ select · Enter open · x stop · Esc/q close',
+    "workflows_footer_run": '↑/↓ select agent · Enter open · x stop · ←/Esc back · q close',
+    "workflows_header_meta": '{running} running · {total} total',
+    "workflows_list_metrics": '{progress} · {agents} agents · {elapsed}',
+    "workflows_no_agents": 'No agent calls yet.',
+    "workflows_phase_unassigned": '(no phase)',
+    "workflows_phase_undeclared": '(undeclared)',
+    "workflows_run_already_finished": 'Run already finished.',
+    "workflows_run_error": 'Error: {error}',
+    "workflows_run_log_path": 'log: {log_path}',
+    "workflows_run_meta": '{run_id} · {status} · {progress} · started {started} · {elapsed}',
+    "workflows_run_missing": 'Run file is gone; returned to the list.',
+    "workflows_run_title": 'Run · {name}',
+    "workflows_section_error": 'Error',
+    "workflows_section_prompt": 'Prompt',
+    "workflows_section_result": 'Result',
+    "workflows_section_tool_calls": 'Tool calls',
+    "workflows_snapshot_empty": 'Workflow runs snapshot: none in this session.',
+    "workflows_snapshot_row": '- {run_id} · {name} · {status} · {progress} · {agents} agents · {elapsed}',
+    "workflows_snapshot_title": 'Workflow runs snapshot:',
+    "workflows_status_aborted": 'aborted',
+    "workflows_status_completed": 'completed',
+    "workflows_status_failed": 'failed',
+    "workflows_status_running": 'running',
+    "workflows_stop_failed": 'Stop failed: {error}',
+    "workflows_stop_requested": 'Stop requested; waiting for the run to settle.',
+    "workflows_stopping": 'stopping…',
+    "workflows_title": 'Workflows',
+    "workflows_tokens_estimate": '~{tokens} tokens',
+    "workflows_tokens_reported": '{tokens} tokens',
+    "workflows_tool_call_line": '{ordinal}. {tool} · {elapsed}',
+    "workflows_tool_call_line_failed": '{ordinal}. {tool} · {elapsed} · failed',
+    "workflows_tool_calls_count": 'tools: {count}',
+    "workflows_tool_calls_none": '(no tool calls)',
+    "workflows_truncated": '… (truncated; full text in {path})',
+    "workflows_unavailable_local_ui": 'Another local UI is open; /workflows is unavailable.',
 }
 
 __all__ = ["EN"]

{comate_cli-0.10.22 → comate_cli-0.11.2}/comate_cli/terminal_agent/i18n/_ja.py

@@ -128,6 +128,20 @@ JA: dict[str, str] = {
         "もう一度 Ctrl+C を押すと終了します。"
     ),
     "mode_switch_failed": "権限モードの切り替えに失敗しました：{error}",
+    "output_style_cancelled": "出力スタイルの変更をキャンセルしました。",
+    "output_style_invalid_argument": "不明な出力スタイル {value} です。default、concise、explanatory、learning、auto、または .md ファイルパスを指定してください。",
+    "output_style_menu_context_current": "現在のスタイル：{name}（ソース：{source}）",
+    "output_style_menu_context_soul_detected": "SOUL.md を検出しました：{path}",
+    "output_style_menu_context_soul_overridden": "SOUL.md（{path}）を検出しましたが、明示的な選択で上書きされています。",
+    "output_style_menu_title": "出力スタイルを選択",
+    "output_style_option_concise_desc": "簡潔型：次の行動が変わる情報だけ",
+    "output_style_option_default_desc": "デフォルトスタイル（結論先行・簡潔）",
+    "output_style_option_explanatory_desc": "解説型：作業しながら設計判断を説明します",
+    "output_style_option_learning_desc": "学習型：小さなコードをあなた自身が書けるように残します",
+    "output_style_option_soul_desc": "SOUL.md に従う（{name}）",
+    "output_style_switch_failed": "出力スタイルの切り替えに失敗しました：{error}",
+    "output_style_switched": "出力スタイル：{previous} → {new}",
+    "output_style_unchanged": "出力スタイルは変わりません（{name}）。",
     "model_menu_effort_auto": "auto の思考量 · /effort で変更",
     "model_menu_effort_hint": "{glyph} {level} の思考量 · /effort で変更",
     "model_menu_effort_unsupported": "{model} は effort 未対応です",
@@ -144,6 +158,13 @@ JA: dict[str, str] = {
     "plugin_auto_update_retry_ok": "更新しました：{plugin_key}",
     "plugin_auto_update_toggled_off": "自動更新を無効にしました：{name}",
     "plugin_auto_update_toggled_on": "自動更新を有効にしました：{name}",
+    "preflight_endpoint_protocol_mismatch": (
+        "{level} のエンドポイント プロトコルが一致していない可能性があります："
+        "provider '{provider}' は {provider_protocol} プロトコルですが、"
+        "base_url {base_url} は {endpoint_protocol} のエンドポイントのようです。"
+        "settings.json の llm_levels と llm_levels_base_url が対応しているか確認してください。"
+    ),
+    "ps_agent_metrics": '{tokens} · ツール呼び出し {tool_calls} 回',
     "ps_unavailable_local_ui": (
         "他のローカル UI が開いているため、/ps は利用できません。"
     ),
@@ -208,6 +229,60 @@ JA: dict[str, str] = {
         "トランスクリプトビューアを開けませんでした：{error}"
     ),
     "webfetch_fetched": "{url} を取得しました ∙ {chars} 文字",
+    "workflow_approval_agent_limit": "agent 呼び出し：最大 {max_calls} 回、並列 {concurrency}；内部 agent は個別承認されません",
+    "workflow_approval_no_phases": "フェーズ：スクリプトに宣言なし",
+    "workflow_approval_phases": "フェーズ：{phases}",
+    "workflow_approval_resume": "{run_id} から再開：完了済みの agent 呼び出しは再生されます",
+    "workflow_approval_script_lines": "スクリプト：{line_count} 行、開始後にディスクへ保存",
+    "workflow_approval_title": "ワークフロー実行",
+    "workflow_summary_resume": "再開 {run_id}",
+    "workflow_summary_script_lines": "{lines} 行のスクリプト",
+    "workflows_agent_metrics": '{model} · {elapsed} · {tokens} · {tool_calls}',
+    "workflows_agent_running": '実行中です（経過 {elapsed}）',
+    "workflows_agent_status_cached": '前回の結果を再利用',
+    "workflows_agent_status_done": '完了',
+    "workflows_agent_status_failed": '失敗',
+    "workflows_agent_status_running": '実行中',
+    "workflows_agent_status_skipped": '未完了',
+    "workflows_agent_title": '{label} の詳細',
+    "workflows_empty": 'このセッションではまだワークフローを実行していません。',
+    "workflows_footer_agent": '↑↓ スクロール · PgUp/PgDn ページ · ← 戻る · q 閉じる',
+    "workflows_footer_list": '↑↓ 移動 · Enter 詳細 · x 停止 · q 閉じる',
+    "workflows_footer_run": '↑↓ agent を選択 · Enter 詳細 · x 停止 · ← 戻る · q 閉じる',
+    "workflows_header_meta": '{running} 件が実行中（全 {total} 件）',
+    "workflows_list_metrics": 'フェーズ {progress} · agent {agents} 件 · 経過 {elapsed}',
+    "workflows_no_agents": 'まだ agent は動いていません。',
+    "workflows_phase_unassigned": 'フェーズなし',
+    "workflows_phase_undeclared": '（途中で追加）',
+    "workflows_run_already_finished": 'この実行はもう終わっています。',
+    "workflows_run_error": 'エラー：{error}',
+    "workflows_run_log_path": 'ログファイル：{log_path}',
+    "workflows_run_meta": '{run_id} · {status} · フェーズ {progress} · {started} 開始 · 経過 {elapsed}',
+    "workflows_run_missing": 'この実行の記録が削除されたため、一覧に戻りました。',
+    "workflows_run_title": '{name} の実行',
+    "workflows_section_error": 'エラー内容',
+    "workflows_section_prompt": 'タスク内容',
+    "workflows_section_result": '結果',
+    "workflows_section_tool_calls": 'ツール呼び出しの記録',
+    "workflows_snapshot_empty": 'ワークフロー実行の一覧：このセッションには記録がありません。',
+    "workflows_snapshot_row": '- {run_id} · {name} · {status} · フェーズ {progress} · agent {agents} 件 · 経過 {elapsed}',
+    "workflows_snapshot_title": 'ワークフロー実行の一覧：',
+    "workflows_status_aborted": '停止済み',
+    "workflows_status_completed": '完了',
+    "workflows_status_failed": '失敗',
+    "workflows_status_running": '実行中',
+    "workflows_stop_failed": '停止できませんでした：{error}',
+    "workflows_stop_requested": '停止を依頼しました。終了処理を待っています。',
+    "workflows_stopping": '停止しています…',
+    "workflows_title": 'ワークフロー',
+    "workflows_tokens_estimate": '約 {tokens} token',
+    "workflows_tokens_reported": '{tokens} token',
+    "workflows_tool_call_line": '{ordinal}. {tool}（{elapsed}）',
+    "workflows_tool_call_line_failed": '{ordinal}. {tool}（{elapsed}）失敗',
+    "workflows_tool_calls_count": 'ツール {count} 回',
+    "workflows_tool_calls_none": '（ツールは使っていません）',
+    "workflows_truncated": '…（長いため省略しました。全文は {path}）',
+    "workflows_unavailable_local_ui": '別のローカル画面が開いています。閉じてから /workflows を使ってください。',
 }
 
 __all__ = ["JA"]

{comate_cli-0.10.22 → comate_cli-0.11.2}/comate_cli/terminal_agent/i18n/_zh.py

@@ -87,6 +87,20 @@ ZH: dict[str, str] = {
         "已发送中断信号。再次按 Esc 可强制中断，再次按 Ctrl+C 可退出。"
     ),
     "mode_switch_failed": "切换权限模式失败：{error}",
+    "output_style_cancelled": "已取消切换输出风格。",
+    "output_style_invalid_argument": "不认识的输出风格 {value}。可选：default、concise、explanatory、learning、auto，或一个 .md 文件路径。",
+    "output_style_menu_context_current": "当前风格：{name}（来源：{source}）",
+    "output_style_menu_context_soul_detected": "检测到 SOUL.md：{path}",
+    "output_style_menu_context_soul_overridden": "检测到 SOUL.md（{path}），但被显式选择覆盖。",
+    "output_style_menu_title": "选择输出风格",
+    "output_style_option_concise_desc": "精简型：只说会改变你下一步的内容",
+    "output_style_option_default_desc": "默认风格（结论先行、保持简洁）",
+    "output_style_option_explanatory_desc": "解释型：边干边讲清楚设计取舍",
+    "output_style_option_learning_desc": "教学型：留小段代码请你亲手完成",
+    "output_style_option_soul_desc": "跟随 SOUL.md（{name}）",
+    "output_style_switch_failed": "切换输出风格失败：{error}",
+    "output_style_switched": "输出风格：{previous} → {new}",
+    "output_style_unchanged": "输出风格没变（{name}）。",
     "model_menu_effort_auto": "auto 思考力度 · 用 /effort 修改",
     "model_menu_effort_hint": "{glyph} {level} 思考力度 · 用 /effort 修改",
     "model_menu_effort_unsupported": "{model} 不支持 effort",
@@ -101,6 +115,12 @@ ZH: dict[str, str] = {
     "plugin_auto_update_retry_ok": "已更新：{plugin_key}",
     "plugin_auto_update_toggled_off": "已关闭自动更新：{name}",
     "plugin_auto_update_toggled_on": "已开启自动更新：{name}",
+    "preflight_endpoint_protocol_mismatch": (
+        "{level} 档的端点协议可能不匹配：provider '{provider}' 走 {provider_protocol} 协议，"
+        "但 base_url {base_url} 看起来是 {endpoint_protocol} 协议的端点。"
+        "请检查 settings.json 里的 llm_levels 与 llm_levels_base_url 是否配套。"
+    ),
+    "ps_agent_metrics": '{tokens} · 工具调用 {tool_calls} 次',
     "ps_unavailable_local_ui": "另一个本地界面正在打开，/ps 暂不可用。",
     "question_no_selection": "无选择",
     "question_unanswered_warning": "还有 {count} 个问题未回答",
@@ -144,6 +164,60 @@ ZH: dict[str, str] = {
     "transcript_empty": "还没有 transcript 条目，请先发送一条消息。",
     "transcript_viewer_failed": "无法打开 transcript 查看器：{error}",
     "webfetch_fetched": "已抓取 {url} ∙ {chars} 字符",
+    "workflow_approval_agent_limit": "agent 调用：最多 {max_calls} 次、并发 {concurrency}；内层 agent 不再逐次审批",
+    "workflow_approval_no_phases": "阶段：脚本未预声明",
+    "workflow_approval_phases": "阶段：{phases}",
+    "workflow_approval_resume": "续跑自 {run_id}：已完成的 agent 调用直接重放",
+    "workflow_approval_script_lines": "脚本：{line_count} 行，落盘后可查看",
+    "workflow_approval_title": "编排工作流",
+    "workflow_summary_resume": "续跑 {run_id}",
+    "workflow_summary_script_lines": "{lines} 行脚本",
+    "workflows_agent_metrics": '{model} · {elapsed} · {tokens} · {tool_calls}',
+    "workflows_agent_running": '正在运行，已用 {elapsed}',
+    "workflows_agent_status_cached": '沿用上次结果',
+    "workflows_agent_status_done": '已完成',
+    "workflows_agent_status_failed": '出错',
+    "workflows_agent_status_running": '进行中',
+    "workflows_agent_status_skipped": '未完成',
+    "workflows_agent_title": '{label} 的详情',
+    "workflows_empty": '本会话还没有运行过工作流。',
+    "workflows_footer_agent": '↑↓ 滚动 · PgUp/PgDn 翻页 · ← 返回 · q 退出',
+    "workflows_footer_list": '↑↓ 移动 · Enter 查看 · x 停止 · q 退出',
+    "workflows_footer_run": '↑↓ 选 agent · Enter 查看 · x 停止 · ← 返回 · q 退出',
+    "workflows_header_meta": '{running} 个正在运行，共 {total} 个',
+    "workflows_list_metrics": '阶段 {progress} · {agents} 个 agent · 用时 {elapsed}',
+    "workflows_no_agents": '还没有任何 agent 开始工作。',
+    "workflows_phase_unassigned": '未分组',
+    "workflows_phase_undeclared": '（临时加入）',
+    "workflows_run_already_finished": '这次运行已经结束了。',
+    "workflows_run_error": '出错：{error}',
+    "workflows_run_log_path": '日志文件：{log_path}',
+    "workflows_run_meta": '{run_id} · {status} · 阶段 {progress} · {started} 开始 · 用时 {elapsed}',
+    "workflows_run_missing": '这次运行的记录已被删除，已回到列表。',
+    "workflows_run_title": '{name} 的运行详情',
+    "workflows_section_error": '出错信息',
+    "workflows_section_prompt": '任务描述',
+    "workflows_section_result": '返回结果',
+    "workflows_section_tool_calls": '工具调用记录',
+    "workflows_snapshot_empty": '工作流运行一览：本会话没有记录。',
+    "workflows_snapshot_row": '- {run_id} · {name} · {status} · 阶段 {progress} · {agents} 个 agent · 用时 {elapsed}',
+    "workflows_snapshot_title": '工作流运行一览：',
+    "workflows_status_aborted": '已停止',
+    "workflows_status_completed": '已完成',
+    "workflows_status_failed": '出错',
+    "workflows_status_running": '进行中',
+    "workflows_stop_failed": '没能停止：{error}',
+    "workflows_stop_requested": '已发出停止请求，正在等它收尾。',
+    "workflows_stopping": '正在停止…',
+    "workflows_title": '工作流',
+    "workflows_tokens_estimate": '约 {tokens} token',
+    "workflows_tokens_reported": '{tokens} token',
+    "workflows_tool_call_line": '{ordinal}. {tool}，用时 {elapsed}',
+    "workflows_tool_call_line_failed": '{ordinal}. {tool}，用时 {elapsed}，出错',
+    "workflows_tool_calls_count": '工具调用 {count} 次',
+    "workflows_tool_calls_none": '（没有调用过工具）',
+    "workflows_truncated": '…（内容过长已截断，完整内容在 {path}）',
+    "workflows_unavailable_local_ui": '另一个本地界面还开着，先关掉它再用 /workflows。',
 }
 
 __all__ = ["ZH"]