multi-claude-cli 0.10.0__tar.gz → 0.11.1__tar.gz

{multi_claude_cli-0.10.0 → multi_claude_cli-0.11.1}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: multi-claude-cli
-Version: 0.10.0
+Version: 0.11.1
 Summary: Run several Claude Code accounts side by side: separate CLAUDE_CONFIG_DIR directories, launchers, proxies, environment variables and arguments.
 License: MIT
 Project-URL: Homepage, https://github.com/jakoes-wu/multi-claude
@@ -333,17 +333,20 @@ old    keychain  token expired  -       inherit                no      -
 
 ### Finding a session
 
-Forgot which account or which directory a session was started in? `sessions` lists the sessions of every account and every project directory, newest first, each followed by the command that resumes it:
+Forgot which account or which directory a session was started in? `sessions` lists the sessions of every account and every project directory, newest first; `--cmd` adds the command that resumes each one:
 
 ```sh
 multi-claude sessions
-# ACCOUNT  UPDATED           CREATED           SESSION ID                            TITLE
-# work     2026-10-11 17:42  2026-10-11 16:05  3c588340-1219-48df-a231-509866789487  Fix the importer
-#   cd /path/to/project && claude-work --resume 3c588340-1219-48df-a231-509866789487
+# TITLE             ACCOUNT  CREATED           UPDATED           SESSION ID
+# Fix the importer  work     2026-10-11 16:05  2026-10-11 17:42  3c588340-1219-48df-a231-509866789487
 multi-claude sessions work            # one account
 multi-claude sessions --grep importer # title, directory or session id contains the text
+multi-claude sessions --grep importer --cmd
+# TITLE             ACCOUNT  CREATED           UPDATED           CMD
+# Fix the importer  work     2026-10-11 16:05  2026-10-11 17:42  cd ~/project && claude-work --resume 3c588340-1219-48df-a231-509866789487
 ```
 
+- `--cmd` puts the resume command in a `CMD` column when the whole table fits the width of the terminal; otherwise each command goes on its own line below its session. The `SESSION ID` column is left out, since each command ends with the session id. The home directory is written as `~`. When the output is piped, the width counts as 80 columns unless `COLUMNS` is set.
 - It shows the 30 most recent sessions; `--limit N` changes that and `--limit 0` shows all. `--json` prints every match with the full fields (`directory`, `created`, `updated`, `title_source`, `file`, `resume`). `conversations` is another name for the same command.
 - The title is the session's name in Claude Code; a session without one shows the beginning of its first prompt, and `-` if it has no prompt at all. Times are local. `UPDATED` is the time of the last message, or the file's modification time when the session has no timestamps.
 - A `~/.claude` that has not been turned into an account is listed as `default`.
@@ -473,7 +476,7 @@ There is no command that switches which account plain `claude` uses. Besides `~/
 
 ### Which account or directory was that session in?
 
-`multi-claude sessions --grep TEXT` searches the titles, directories and ids of the sessions of all accounts and prints the command that resumes each match. See [Finding a session](#finding-a-session).
+`multi-claude sessions --grep TEXT` searches the titles, directories and ids of the sessions of all accounts; add `--cmd` to print the command that resumes each match. See [Finding a session](#finding-a-session).
 
 ### How do I continue a conversation with another account?
 
@@ -522,7 +525,7 @@ It never copies, exports, uploads or backs up a login. One command touches one:
 | `multi-claude path NAME` | Print the account's directory. |
 | `multi-claude code NAME [PATH] [-- ARGS ...]` | Open a separate VS Code instance with the account's environment (experimental). |
 | `multi-claude handoff TARGET [--from NAME] [--session ID] [--force]` | Copy a session to another account and print the command that resumes it there. |
-| `multi-claude sessions [NAME] [--grep TEXT] [--limit N] [--json]` | List the sessions of every account and directory, newest first, with the command that resumes each. Also `conversations`. |
+| `multi-claude sessions [NAME] [--grep TEXT] [--limit N] [--cmd] [--json]` | List the sessions of every account and directory, newest first; `--cmd` adds the command that resumes each. Also `conversations`. |
 | `multi-claude statusline install\|uninstall FILE` | Wrap the statusLine command in a settings file so that `usage` gets fresh numbers, or restore it. |
 | `multi-claude completion bash\|zsh\|fish` | Print a shell completion script. |
 

{multi_claude_cli-0.10.0 → multi_claude_cli-0.11.1}/README.md

@@ -306,17 +306,20 @@ old    keychain  token expired  -       inherit                no      -
 
 ### Finding a session
 
-Forgot which account or which directory a session was started in? `sessions` lists the sessions of every account and every project directory, newest first, each followed by the command that resumes it:
+Forgot which account or which directory a session was started in? `sessions` lists the sessions of every account and every project directory, newest first; `--cmd` adds the command that resumes each one:
 
 ```sh
 multi-claude sessions
-# ACCOUNT  UPDATED           CREATED           SESSION ID                            TITLE
-# work     2026-10-11 17:42  2026-10-11 16:05  3c588340-1219-48df-a231-509866789487  Fix the importer
-#   cd /path/to/project && claude-work --resume 3c588340-1219-48df-a231-509866789487
+# TITLE             ACCOUNT  CREATED           UPDATED           SESSION ID
+# Fix the importer  work     2026-10-11 16:05  2026-10-11 17:42  3c588340-1219-48df-a231-509866789487
 multi-claude sessions work            # one account
 multi-claude sessions --grep importer # title, directory or session id contains the text
+multi-claude sessions --grep importer --cmd
+# TITLE             ACCOUNT  CREATED           UPDATED           CMD
+# Fix the importer  work     2026-10-11 16:05  2026-10-11 17:42  cd ~/project && claude-work --resume 3c588340-1219-48df-a231-509866789487
 ```
 
+- `--cmd` puts the resume command in a `CMD` column when the whole table fits the width of the terminal; otherwise each command goes on its own line below its session. The `SESSION ID` column is left out, since each command ends with the session id. The home directory is written as `~`. When the output is piped, the width counts as 80 columns unless `COLUMNS` is set.
 - It shows the 30 most recent sessions; `--limit N` changes that and `--limit 0` shows all. `--json` prints every match with the full fields (`directory`, `created`, `updated`, `title_source`, `file`, `resume`). `conversations` is another name for the same command.
 - The title is the session's name in Claude Code; a session without one shows the beginning of its first prompt, and `-` if it has no prompt at all. Times are local. `UPDATED` is the time of the last message, or the file's modification time when the session has no timestamps.
 - A `~/.claude` that has not been turned into an account is listed as `default`.
@@ -446,7 +449,7 @@ There is no command that switches which account plain `claude` uses. Besides `~/
 
 ### Which account or directory was that session in?
 
-`multi-claude sessions --grep TEXT` searches the titles, directories and ids of the sessions of all accounts and prints the command that resumes each match. See [Finding a session](#finding-a-session).
+`multi-claude sessions --grep TEXT` searches the titles, directories and ids of the sessions of all accounts; add `--cmd` to print the command that resumes each match. See [Finding a session](#finding-a-session).
 
 ### How do I continue a conversation with another account?
 
@@ -495,7 +498,7 @@ It never copies, exports, uploads or backs up a login. One command touches one:
 | `multi-claude path NAME` | Print the account's directory. |
 | `multi-claude code NAME [PATH] [-- ARGS ...]` | Open a separate VS Code instance with the account's environment (experimental). |
 | `multi-claude handoff TARGET [--from NAME] [--session ID] [--force]` | Copy a session to another account and print the command that resumes it there. |
-| `multi-claude sessions [NAME] [--grep TEXT] [--limit N] [--json]` | List the sessions of every account and directory, newest first, with the command that resumes each. Also `conversations`. |
+| `multi-claude sessions [NAME] [--grep TEXT] [--limit N] [--cmd] [--json]` | List the sessions of every account and directory, newest first; `--cmd` adds the command that resumes each. Also `conversations`. |
 | `multi-claude statusline install\|uninstall FILE` | Wrap the statusLine command in a settings file so that `usage` gets fresh numbers, or restore it. |
 | `multi-claude completion bash\|zsh\|fish` | Print a shell completion script. |
 

{multi_claude_cli-0.10.0 → multi_claude_cli-0.11.1}/src/multi_claude/__init__.py

@@ -1,3 +1,3 @@
 """multi-claude：管理多个 Claude Code 账号的配置目录、启动命令、代理与启动参数。"""
 
-__version__ = "0.10.0"
+__version__ = "0.11.1"

{multi_claude_cli-0.10.0 → multi_claude_cli-0.11.1}/src/multi_claude/cli.py

@@ -15,6 +15,7 @@ import shutil
 import stat
 import subprocess
 import sys
+import unicodedata
 from datetime import datetime, timedelta, timezone
 from typing import List, NamedTuple, Optional, Tuple
 
@@ -279,15 +280,19 @@ def build_parser() -> argparse.ArgumentParser:
     _add_dry_run(p_handoff)
 
     p_sessions = sub.add_parser("sessions", description="list the Claude sessions of every account and every directory, "
-                                                 "newest first, each with the command that resumes it "
+                                                 "newest first; --cmd adds the command that resumes each "
                                                  "(read-only; also available as `conversations`): "
-                                                 "sessions [NAME] [--grep TEXT] [--limit N] [--json]")
+                                                 "sessions [NAME] [--grep TEXT] [--limit N] [--cmd] [--json]")
     p_sessions.add_argument("name", nargs="?", metavar="NAME",
                             help="only this account (`default` is the unregistered ~/.claude)")
     p_sessions.add_argument("--grep", metavar="TEXT",
                             help="only sessions whose title, directory or id contains TEXT (case-insensitive)")
     p_sessions.add_argument("--limit", type=int, metavar="N",
                             help="show at most N sessions, 0 for all (default: 30, or all with --json)")
+    p_sessions.add_argument("--cmd", action="store_true",
+                            help="also print the command that resumes each session (a CMD column when the "
+                                 "terminal is wide enough, otherwise a line below each session); the SESSION ID "
+                                 "column is then left out, since the command contains the id")
     p_sessions.add_argument("--json", action="store_true", help="print machine-readable JSON")
     return parser
 
@@ -427,7 +432,7 @@ def main(argv: Optional[List[str]] = None) -> int:
             return cmd_usage(args.name, args.json)
         if args.command == "sessions":
             # 只读：不加写锁，迁移未完成时也放行（与 list、usage 相同）。
-            return cmd_sessions(args.name, args.grep, args.limit, args.json)
+            return cmd_sessions(args.name, args.grep, args.limit, args.json, args.cmd)
         if args.command == "which":
             return cmd_which(args.path)
         if args.command == "statusline":
@@ -1767,8 +1772,11 @@ def _session_sources(config: Config, exists: bool) -> List[_SessionSource]:
     return sources
 
 
-def cmd_sessions(name: Optional[str], grep: Optional[str], limit: Optional[int], as_json: bool) -> int:
-    """只读列出各账号、各目录的会话，最近更新的在前，每条附续聊命令（方案 feature-session-list）。
+def cmd_sessions(name: Optional[str], grep: Optional[str], limit: Optional[int], as_json: bool,
+                 show_cmd: bool = False) -> int:
+    """只读列出各账号、各目录的会话，最近更新的在前（方案 feature-session-list、feature-sessions-cmd-column）。
+
+    show_cmd 只影响文本输出：为真时才打印续聊命令。JSON 输出始终带 resume 字段，不受它影响。
 
     会把会话标题或首条提问的开头打印出来（这是对话内容的片段），不保存、不联网、不读凭据。
     同一会话 ID 在多个账号里各有一份时（handoff 复制过）各列一行，不合并。
@@ -1821,31 +1829,95 @@ def cmd_sessions(name: Optional[str], grep: Optional[str], limit: Optional[int],
     if not entries:
         print("no sessions found")
         return accounts.EXIT_OK
-    # 不用 _print_table：每条会话下面要插一行续聊命令。标题放最后一列，中文标题的显示宽度不影响前面各列对齐。
-    rows = [("ACCOUNT", "UPDATED", "CREATED", "SESSION ID", "TITLE", None)]
+    # 不用 _print_table：标题在第一列且可能是中文，要按终端显示宽度对齐；带 --cmd 时还要在两种排版之间选。
+    header = ["TITLE", "ACCOUNT", "CREATED", "UPDATED", "SESSION ID"]
+    rows = []
+    commands = []
     for source, session, full_title in entries:
-        resume = _resume_command(source, session)
+        rows.append([sessions.clean_title(full_title, _TITLE_TEXT_CHARS) or "-", source.name,
+                     _local_minute(session.created), _local_minute(session.updated), session.session_id])
+        resume = _resume_command(source, session, tilde=True)
         if session.directory is None:
             resume += "  (start directory unknown; project {})".format(
                 sessions.clean_title(session.project, len(session.project)))
-        rows.append((source.name, _local_minute(session.updated), _local_minute(session.created), session.session_id,
-                     sessions.clean_title(full_title, _TITLE_TEXT_CHARS) or "-", resume))
-    widths = [max(len(row[index]) for row in rows) for index in range(4)]
-    for row in rows:
-        print("  ".join([cell.ljust(width) for cell, width in zip(row[:4], widths)] + [row[4]]).rstrip())
-        if row[5] is not None:
-            print("  " + row[5])
+        commands.append(resume)
+    if not show_cmd:
+        for line in _aligned([header] + rows):
+            print(line)
+    else:
+        # 续聊命令里总有 --resume <会话ID>（目录未知时也有），带 --cmd 时两种排版都去掉 SESSION ID 列，
+        # 免得同一串 ID 在一条会话里出现两次；不带 --cmd 时它是唯一能看到 ID 的地方，必须保留。
+        header = header[:-1]
+        rows = [row[:-1] for row in rows]
+        # 续聊命令很长：整张表（表头与数据行，不含末尾的截断提示）放得下终端宽度才并成 CMD 列，
+        # 否则命令在窄终端里会从中间折行，比单独一行更难读、也更难复制。按整张表决定，不逐行切换。
+        wide = _aligned([header + ["CMD"]] + [row + [command] for row, command in zip(rows, commands)])
+        if max(_display_width(line) for line in wide) <= shutil.get_terminal_size((80, 24)).columns:
+            for line in wide:
+                print(line)
+        else:
+            narrow = _aligned([header] + rows)
+            print(narrow[0])
+            for line, command in zip(narrow[1:], commands):
+                print(line)
+                print("  " + command)
     if total > len(entries):
         print("showing {} of {} sessions; use --limit 0 for all, or --grep TEXT to narrow".format(len(entries), total))
     return accounts.EXIT_OK
 
 
-def _resume_command(source: _SessionSource, session: sessions.SessionInfo) -> str:
-    """续聊命令。Claude 按当前目录找会话，所以先 cd 到会话启动时的目录；目录未知时只给后半段。"""
+def _resume_command(source: _SessionSource, session: sessions.SessionInfo, tilde: bool = False) -> str:
+    """续聊命令。Claude 按当前目录找会话，所以先 cd 到会话启动时的目录；目录未知时只给后半段。
+
+    tilde 为真时把主目录写成 ~，只用于给人看的文本输出；JSON 的 resume 字段保持绝对路径。
+    """
     resume = "{} --resume {}".format(source.launcher, session.session_id)
     if session.directory is None:
         return resume
-    return "cd {} && {}".format(shlex.quote(session.directory), resume)
+    target = _tilde_path(session.directory) if tilde else shlex.quote(session.directory)
+    return "cd {} && {}".format(target, resume)
+
+
+def _tilde_path(directory: str) -> str:
+    """把主目录之下的目录写成 ~/…，可直接拼在 cd 后面；不在主目录之下时与 shlex.quote 相同。
+
+    按字符串前缀比较，不解析软链：会话里记的是当时的目录原文。
+    """
+    home = os.path.expanduser("~")
+    if not os.path.isabs(home) or home == os.sep:
+        return shlex.quote(directory)
+    if directory == home:
+        return "~"
+    if not directory.startswith(home + os.sep):
+        return shlex.quote(directory)
+    rest = directory[len(home) + 1:]
+    # 波浪号必须留在引号外面，放进引号里 shell 不会把它展开成主目录
+    return "~/" + shlex.quote(rest) if rest else "~"
+
+
+def _display_width(text: str) -> int:
+    """文字在终端里占的格数：中日韩等宽字符占 2 格，组合字符和零宽连接符不占格。
+
+    表情符号的组合序列在不同终端里宽度不一，这里可能差一格。
+    """
+    width = 0
+    for char in text:
+        # U+200C、U+200D 是 clean_title 特意保留的零宽字符
+        if char in "\u200c\u200d" or unicodedata.category(char) in ("Mn", "Me"):
+            continue
+        width += 2 if unicodedata.east_asian_width(char) in ("W", "F") else 1
+    return width
+
+
+def _aligned(rows: List[List[str]]) -> List[str]:
+    """按显示宽度对齐各列，返回每行的文字；列间两个空格，最后一列不补空格。"""
+    widths = [max(_display_width(row[index]) for row in rows) for index in range(len(rows[0]))]
+    lines = []
+    for row in rows:
+        # str.ljust 按字符数补空格，中文标题会少补；这里按显示宽度补
+        cells = [cell + " " * (width - _display_width(cell)) for cell, width in zip(row, widths)]
+        lines.append("  ".join(cells).rstrip())
+    return lines
 
 
 def _local_minute(moment: Optional[datetime]) -> str:

{multi_claude_cli-0.10.0 → multi_claude_cli-0.11.1}/src/multi_claude_cli.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: multi-claude-cli
-Version: 0.10.0
+Version: 0.11.1
 Summary: Run several Claude Code accounts side by side: separate CLAUDE_CONFIG_DIR directories, launchers, proxies, environment variables and arguments.
 License: MIT
 Project-URL: Homepage, https://github.com/jakoes-wu/multi-claude
@@ -333,17 +333,20 @@ old    keychain  token expired  -       inherit                no      -
 
 ### Finding a session
 
-Forgot which account or which directory a session was started in? `sessions` lists the sessions of every account and every project directory, newest first, each followed by the command that resumes it:
+Forgot which account or which directory a session was started in? `sessions` lists the sessions of every account and every project directory, newest first; `--cmd` adds the command that resumes each one:
 
 ```sh
 multi-claude sessions
-# ACCOUNT  UPDATED           CREATED           SESSION ID                            TITLE
-# work     2026-10-11 17:42  2026-10-11 16:05  3c588340-1219-48df-a231-509866789487  Fix the importer
-#   cd /path/to/project && claude-work --resume 3c588340-1219-48df-a231-509866789487
+# TITLE             ACCOUNT  CREATED           UPDATED           SESSION ID
+# Fix the importer  work     2026-10-11 16:05  2026-10-11 17:42  3c588340-1219-48df-a231-509866789487
 multi-claude sessions work            # one account
 multi-claude sessions --grep importer # title, directory or session id contains the text
+multi-claude sessions --grep importer --cmd
+# TITLE             ACCOUNT  CREATED           UPDATED           CMD
+# Fix the importer  work     2026-10-11 16:05  2026-10-11 17:42  cd ~/project && claude-work --resume 3c588340-1219-48df-a231-509866789487
 ```
 
+- `--cmd` puts the resume command in a `CMD` column when the whole table fits the width of the terminal; otherwise each command goes on its own line below its session. The `SESSION ID` column is left out, since each command ends with the session id. The home directory is written as `~`. When the output is piped, the width counts as 80 columns unless `COLUMNS` is set.
 - It shows the 30 most recent sessions; `--limit N` changes that and `--limit 0` shows all. `--json` prints every match with the full fields (`directory`, `created`, `updated`, `title_source`, `file`, `resume`). `conversations` is another name for the same command.
 - The title is the session's name in Claude Code; a session without one shows the beginning of its first prompt, and `-` if it has no prompt at all. Times are local. `UPDATED` is the time of the last message, or the file's modification time when the session has no timestamps.
 - A `~/.claude` that has not been turned into an account is listed as `default`.
@@ -473,7 +476,7 @@ There is no command that switches which account plain `claude` uses. Besides `~/
 
 ### Which account or directory was that session in?
 
-`multi-claude sessions --grep TEXT` searches the titles, directories and ids of the sessions of all accounts and prints the command that resumes each match. See [Finding a session](#finding-a-session).
+`multi-claude sessions --grep TEXT` searches the titles, directories and ids of the sessions of all accounts; add `--cmd` to print the command that resumes each match. See [Finding a session](#finding-a-session).
 
 ### How do I continue a conversation with another account?
 
@@ -522,7 +525,7 @@ It never copies, exports, uploads or backs up a login. One command touches one:
 | `multi-claude path NAME` | Print the account's directory. |
 | `multi-claude code NAME [PATH] [-- ARGS ...]` | Open a separate VS Code instance with the account's environment (experimental). |
 | `multi-claude handoff TARGET [--from NAME] [--session ID] [--force]` | Copy a session to another account and print the command that resumes it there. |
-| `multi-claude sessions [NAME] [--grep TEXT] [--limit N] [--json]` | List the sessions of every account and directory, newest first, with the command that resumes each. Also `conversations`. |
+| `multi-claude sessions [NAME] [--grep TEXT] [--limit N] [--cmd] [--json]` | List the sessions of every account and directory, newest first; `--cmd` adds the command that resumes each. Also `conversations`. |
 | `multi-claude statusline install\|uninstall FILE` | Wrap the statusLine command in a settings file so that `usage` gets fresh numbers, or restore it. |
 | `multi-claude completion bash\|zsh\|fish` | Print a shell completion script. |
 

{multi_claude_cli-0.10.0 → multi_claude_cli-0.11.1}/tests/test_sessions.py

@@ -5,6 +5,8 @@
 
 import json
 import os
+import shlex
+import shutil
 import sys
 import unittest
 
@@ -83,15 +85,69 @@ class ListingTest(SessionsTestCase):
         self.make(self.dir_b, SID3, [user("q", cwd="/w/third", ts="2026-10-02T00:00:00.000Z"), ai("Middle session")],
                   project="-w-third")
         out = self.ok("sessions", env={"TZ": "Asia/Shanghai"}).out.splitlines()
-        self.assertEqual(out[0].split(), ["ACCOUNT", "UPDATED", "CREATED", "SESSION", "ID", "TITLE"])
-        # 按最后更新时间从新到旧；时间按本地时区（UTC+8）显示
-        self.assertEqual(out[1], "a        2026-10-03 09:00  2026-10-03 00:30  {}  Newest session".format(SID2))
-        self.assertEqual(out[2], "  cd /w/other && claude-a --resume " + SID2)
-        self.assertEqual(out[3], "b        2026-10-02 08:00  2026-10-02 08:00  {}  Middle session".format(SID3))
-        self.assertEqual(out[4], "  cd /w/third && claude-b --resume " + SID3)
-        self.assertEqual(out[5], "a        2026-10-01 11:30  2026-10-01 10:00  {}  Older session".format(SID1))
-        self.assertEqual(out[6], "  cd /w/proj && claude-a --resume " + SID1)
-        self.assertEqual(len(out), 7)
+        # 按最后更新时间从新到旧；时间按本地时区（UTC+8）显示。不带 --cmd 时每条会话一行，不打印续聊命令
+        self.assertEqual(out, [
+            "TITLE           ACCOUNT  CREATED           UPDATED           SESSION ID",
+            "Newest session  a        2026-10-03 00:30  2026-10-03 09:00  " + SID2,
+            "Middle session  b        2026-10-02 08:00  2026-10-02 08:00  " + SID3,
+            "Older session   a        2026-10-01 10:00  2026-10-01 11:30  " + SID1])
+
+    def test_cmd_column_when_the_terminal_is_wide_enough(self):
+        self.make(self.dir_a, SID1, [user("q"), ai("One")])
+        self.make(self.dir_b, SID2, [user("q", cwd="/w/other", ts="2026-10-01T01:00:00.000Z"), ai("Two")],
+                  project="-w-other")
+        env = {"TZ": "Asia/Shanghai", "COLUMNS": "300"}
+        wide = self.ok("sessions", "--cmd", env=env).out.splitlines()
+        self.assertEqual(wide, [
+            "TITLE  ACCOUNT  CREATED           UPDATED           CMD",
+            "One    a        2026-10-01 10:00  2026-10-01 10:00  cd /w/proj && claude-a --resume " + SID1,
+            "Two    b        2026-10-01 09:00  2026-10-01 09:00  cd /w/other && claude-b --resume " + SID2])
+        # 会话 ID 已在命令末尾，带 --cmd 时不再单列一次
+        # 放不下时整张表改成命令单独一行；恰好放得下仍算放得下。标题与目录都是 ASCII，len 就是显示宽度
+        widest = max(len(line) for line in wide)
+        self.assertEqual(self.ok("sessions", "--cmd", env=dict(env, COLUMNS=str(widest))).out.splitlines(), wide)
+        narrow = [
+            "TITLE  ACCOUNT  CREATED           UPDATED",
+            "One    a        2026-10-01 10:00  2026-10-01 10:00",
+            "  cd /w/proj && claude-a --resume " + SID1,
+            "Two    b        2026-10-01 09:00  2026-10-01 09:00",
+            "  cd /w/other && claude-b --resume " + SID2]
+        self.assertEqual(self.ok("sessions", "--cmd", env=dict(env, COLUMNS=str(widest - 1))).out.splitlines(), narrow)
+        self.assertEqual(self.ok("sessions", "--cmd", env=dict(env, COLUMNS="40")).out.splitlines(), narrow)
+
+    def test_wide_titles_keep_the_columns_aligned(self):
+        # 中文一个字占两格：按字符数补空格的话，中文标题后面的列会往右错
+        self.make(self.dir_a, SID1, [user("q"), custom("修复解析")])
+        self.make(self.dir_a, SID2, [user("q", ts="2026-10-01T01:00:00.000Z"), custom("abcdefgh")])
+        tail = "  a        2026-10-01 {}:00  2026-10-01 {}:00  {}"
+        env = {"TZ": "Asia/Shanghai"}
+        self.assertEqual(self.ok("sessions", env=env).out.splitlines(),
+                         ["TITLE     ACCOUNT  CREATED           UPDATED           SESSION ID",
+                          "修复解析" + tail.format("10", "10", SID1), "abcdefgh" + tail.format("09", "09", SID2)])
+        self.make(self.dir_a, SID3, [user("q", ts="2026-10-01T00:00:00.000Z"), custom("修复解析器")])
+        self.assertEqual(self.ok("sessions", env=env).out.splitlines(),
+                         ["TITLE       ACCOUNT  CREATED           UPDATED           SESSION ID",
+                          "修复解析  " + tail.format("10", "10", SID1), "abcdefgh  " + tail.format("09", "09", SID2),
+                          "修复解析器" + tail.format("08", "08", SID3)])
+        # 零宽连接符与组合字符不占格：e 加组合重音符是 1 格，这个标题共 8 格，和 abcdefgh 一样宽
+        self.make(self.dir_a, SID3, [user("q", ts="2026-10-01T00:00:00.000Z"), custom("abcde\u0301f\u200dgh")])
+        self.assertEqual(self.ok("sessions", env=env).out.splitlines()[1:],
+                         ["修复解析" + tail.format("10", "10", SID1), "abcdefgh" + tail.format("09", "09", SID2),
+                          "abcde\u0301f\u200dgh" + tail.format("08", "08", SID3)])
+
+    def test_cmd_writes_the_home_directory_as_a_tilde(self):
+        cases = ((os.path.join(self.home, "sub"), "cd ~/sub"), (self.home, "cd ~"), (self.home + "/", "cd ~"),
+                 (os.path.join(self.home, "my proj"), "cd ~/'my proj'"), ("/w/proj", "cd /w/proj"),
+                 (self.home + "2", "cd " + shlex.quote(self.home + "2")))
+        for cwd, expected in cases:
+            self.make(self.dir_a, SID1, [user("q", cwd=cwd), ai("T")], project=sessions.encode(cwd))
+            line = "  {} && claude-a --resume {}".format(expected, SID1)
+            self.assertEqual(self.ok("sessions", "--cmd").out.splitlines()[2], line, cwd)
+            # JSON 给脚本用，保持绝对路径；--cmd 对它没有影响
+            data = self.listing()
+            self.assertTrue(data["sessions"][0]["resume"].startswith("cd " + shlex.quote(cwd) + " && "), cwd)
+            self.assertEqual(self.listing("--cmd"), data)
+            shutil.rmtree(os.path.join(self.dir_a, "projects"))
 
     def test_t16_json_structure(self):
         path = self.make(self.dir_a, SID1, [user("q"), assistant("2026-10-01T03:30:00.000Z"), custom("Named")])
@@ -147,7 +203,7 @@ class TitleTest(SessionsTestCase):
         self.make(self.dir_a, SID1, [{"type": "mode", "mode": "x"}, {"type": "system", "cwd": "/w/proj",
                                                                     "timestamp": "2026-10-01T02:00:00.000Z"}])
         self.assertEqual((self.one()["title"], self.one()["title_source"]), ("", "none"))
-        self.assertTrue(self.ok("sessions").out.splitlines()[1].endswith(SID1 + "  -"))
+        self.assertTrue(self.ok("sessions").out.splitlines()[1].startswith("-      a  "))
 
     def test_t3_first_plain_prompt(self):
         self.make(self.dir_a, SID1, [
@@ -170,8 +226,9 @@ class TitleTest(SessionsTestCase):
         out = self.ok("sessions").out
         self.assertNotIn("\x1b", out)
         self.assertNotIn("\x07", out)
-        self.assertEqual(len(out.splitlines()), 3)
-        shown = out.splitlines()[1].split(SID1 + "  ", 1)[1]
+        self.assertEqual(len(out.splitlines()), 2)
+        # 标题在第一列；清洗后的标题里没有连续空格，按列间的两个空格切开即可
+        shown = out.splitlines()[1].split("  a  ", 1)[0]
         self.assertEqual(len(shown), 61)
         self.assertTrue(shown.endswith("…"))
 
@@ -193,7 +250,7 @@ class TitleTest(SessionsTestCase):
         self.make(self.dir_a, SID1, [user("q", cwd=cwd), custom(title)], project=sessions.encode(cwd))
         row = self.one()
         self.assertEqual((row["title"], row["directory"]), (title, cwd))
-        self.assertIn("cd '{}' && claude-a --resume {}".format(cwd, SID1), self.ok("sessions").out)
+        self.assertIn("cd '{}' && claude-a --resume {}".format(cwd, SID1), self.ok("sessions", "--cmd").out)
 
     def test_t6_title_only_near_the_start(self):
         # 标题行在首条提问之前、又被挤出尾部窗口：靠头部扫描取到
@@ -231,7 +288,8 @@ class FieldFallbackTest(SessionsTestCase):
     def test_t5c_cwd_with_a_newline_is_not_used(self):
         self.make(self.dir_a, SID1, [user("q", cwd="/w/pr\noj"), assistant("2026-10-01T05:00:00.000Z"), ai("T")])
         self.assertIsNone(self.one()["directory"])
-        self.assertEqual(len(self.ok("sessions").out.splitlines()), 3)
+        self.assertEqual(len(self.ok("sessions").out.splitlines()), 2)
+        self.assertEqual(len(self.ok("sessions", "--cmd").out.splitlines()), 3)
 
     def test_t7_missing_fields_bad_lines_and_half_line(self):
         self.make(self.dir_a, SID1, [b"not json\n", b"[1, 2]\n", {"type": "mode", "mode": "x"}, ai("Only a title")],
@@ -242,9 +300,13 @@ class FieldFallbackTest(SessionsTestCase):
         self.assertEqual(row["created"], None)
         self.assertEqual(row["updated"], "2026-09-21T14:13:20Z")
         self.assertEqual(row["resume"], "claude-a --resume " + SID1)
-        out = self.ok("sessions").out.splitlines()
-        self.assertEqual(out[2], "  claude-a --resume {}  (start directory unknown; project -w-proj)".format(SID1))
+        unknown = "claude-a --resume {}  (start directory unknown; project -w-proj)".format(SID1)
+        out = self.ok("sessions", "--cmd", env={"COLUMNS": "40"}).out.splitlines()
+        self.assertEqual(out[2], "  " + unknown)
         self.assertIn("  -  ", out[1])
+        self.assertTrue(self.ok("sessions", "--cmd", env={"COLUMNS": "300"}).out.splitlines()[1].endswith("  " + unknown))
+        # 不带 --cmd 时没有命令，这段说明也不出现
+        self.assertNotIn("start directory unknown", self.ok("sessions").out)
 
     def test_t7b_deeply_nested_line_is_skipped(self):
         # 嵌套太深的行让 json 抛 RecursionError（不是 ValueError）；它只是一条坏行，别的字段照常取
@@ -260,9 +322,9 @@ class FieldFallbackTest(SessionsTestCase):
         for zone, sid in (("America/New_York", SID1), ("Asia/Shanghai", SID2), ("UTC", None)):
             result = self.ok("sessions", env={"TZ": zone})
             self.assertNotIn("Traceback", result.err)
-            self.assertEqual(len(result.out.splitlines()), 5, result)
+            self.assertEqual(len(result.out.splitlines()), 3, result)
             # 换算不了的时间显示 -（这两条会话的创建与更新时间是同一个值）；UTC 下两条都能显示
-            dashed = [line.split()[3] for line in result.out.splitlines() if line.split()[1:3] == ["-", "-"]]
+            dashed = [line.split()[4] for line in result.out.splitlines() if line.split()[2:4] == ["-", "-"]]
             self.assertEqual(dashed, [sid] if sid else [], result)
         self.assertEqual(len(self.listing()["sessions"]), 2)
 
@@ -375,9 +437,10 @@ class FilterTest(SessionsTestCase):
         data = self.listing("--limit", "1")
         self.assertEqual(([row["session_id"] for row in data["sessions"]], data["total"]), ([SID2], 3))
         out = self.ok("sessions", "--limit", "1").out.splitlines()
-        self.assertEqual(len(out), 4)
+        self.assertEqual(len(out), 3)
         self.assertEqual(out[-1], "showing 1 of 3 sessions; use --limit 0 for all, or --grep TEXT to narrow")
-        self.assertEqual(len(self.ok("sessions", "--limit", "0").out.splitlines()), 7)
+        self.assertEqual(len(self.ok("sessions", "--limit", "0").out.splitlines()), 4)
+        self.assertEqual(len(self.ok("sessions", "--limit", "0", "--cmd").out.splitlines()), 7)
         result = self.run_cli("sessions", "--limit", "-1")
         self.assertEqual(result.code, 2)
         self.assertIn("--limit must be 0 or more", result.err)
@@ -387,7 +450,7 @@ class FilterTest(SessionsTestCase):
         for index in range(28):
             self.make(self.dir_b, "44444444-aaaa-bbbb-cccc-0000000000{:02d}".format(index), [user("q")])
         out = self.ok("sessions").out.splitlines()
-        self.assertEqual(len(out), 1 + 30 * 2 + 1)
+        self.assertEqual(len(out), 1 + 30 + 1)
         self.assertEqual(out[-1], "showing 30 of 31 sessions; use --limit 0 for all, or --grep TEXT to narrow")
         self.assertEqual(len(self.listing()["sessions"]), 31)