@kadaliao/dsh-bots 0.8.0 → 0.9.0

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
package/README.md CHANGED
@@ -6,7 +6,8 @@ Grok-style bots inside DeepSeek Harness: a team of bots that work together. Each
6
6
  persona, model, skills, abilities, apps, scheduled tasks and memory**, and its own conversation. Bots **message each
7
7
  other**, and one of them, the **dispatcher**, splits the work, hands it out, follows up and reports back. Each piece
8
8
  of work handed out is a **task** that the bot works on in a conversation of its own, so one bot can work on several
9
- at once.
9
+ at once. You chat with a bot as in a Feishu group: send messages whenever you like, and each one gets its answers in a
10
+ **topic** of its own.
10
11
 
11
12
  It is built for two situations:
12
13
 
@@ -61,6 +62,30 @@ Bots answer in the language you write in, so the same team works in Chinese and
61
62
  Each message from Atlas to a teammate opens a task, the three tasks run at the same time, and Forge's message is the
62
63
  report on its task.
63
64
 
65
+ ## The chat: one topic per request
66
+
67
+ Talk to a bot as you would in a Feishu group: send as many messages as you like, whenever you like, and each one gets
68
+ its answers in a thread of its own.
69
+
70
+ <!-- screenshot: chat with topics -->
71
+
72
+ - **Every message is a topic.** What you send shows at once, marked **Queued** until the bot takes it. The answers to
73
+ it, and everything said about the work it started, collect under it. A bar under the message shows how many replies
74
+ there are, the latest one, the task working on it (with its live status) and a dot for anything you haven't read;
75
+ click it to open the topic beside the chat. When a topic out of sight gets a reply, **N topics have new replies**
76
+ shows above the box and takes you there.
77
+ - **Reply to any message.** Hover over a message and click **Reply**: what you write goes into that message's topic.
78
+ In a topic whose task is still open (or finished in the last day), it goes straight to the bot doing the task; the
79
+ dispatcher only gets a note, which doesn't wake it.
80
+ - **The dispatcher answers every message within a step or two**, because it doesn't do the work in its own
81
+ conversation: it hands it to the teammate that fits, or to itself as a task of its own when none does, and is free
82
+ again at once. Two requests sent back to back run side by side, and neither is dropped for the other. A message that
83
+ continues something already under way is added to that task and filed under its topic (**Filed under "…"**).
84
+ - **The bot doing the work answers you itself**, in the topic the request came from, with the whole result. Only when
85
+ the dispatcher splits one request between several bots and has to put the results together does it ask for their
86
+ reports back and answer you itself.
87
+ - Messages from before 0.9 stay in the main timeline as they were.
88
+
64
89
  ## Tasks: delegated work in conversations of their own
65
90
 
66
91
  When a bot hands work to a teammate, the work becomes a **task**. The teammate works on each task in a conversation
@@ -69,10 +94,12 @@ for you. You don't open tasks yourself: you ask the dispatcher (or any bot), and
69
94
 
70
95
  <!-- screenshot: tasks view -->
71
96
 
72
- - **Reports**: the bot doing a task reports to the bot that asked, with a status: **done** with the result,
73
- **question** when it needs an answer, **blocked** when it cannot go on, or **progress**, which is kept on the task
74
- without interrupting anyone. Done, question and blocked reach the bot that asked at once, and it passes them on
75
- (the dispatcher, to you). A report from a task that has already ended, or from a sub-task whose task has ended, is
97
+ - **Reports**: the bot doing a task reports with a status: **done** with the result, **question** when it needs an
98
+ answer, **blocked** when it cannot go on, or **progress**, which is kept on the task without interrupting anyone.
99
+ Done, question and blocked go to you as they are, in the topic or chat the request came from, and the bot that asked
100
+ gets a note that doesn't wake it. When the bot that asked wants the reports back (to put several results together),
101
+ they reach it at once instead and it passes them on; so do a sub-task's reports, and those of a task that came from
102
+ no chat (a scheduled run). A report from a task that has already ended, or from a sub-task whose task has ended, is
76
103
  only kept on the task. Anything else the working bot writes stays in the task's conversation.
77
104
  - **Sub-tasks**: a bot working on a task can ask a teammate for help; that opens a sub-task, whose report comes back
78
105
  to the task that asked for it.
@@ -90,7 +117,8 @@ for you. You don't open tasks yourself: you ask the dispatcher (or any bot), and
90
117
  **Mark done** and **Stop** both end the task and stop its bot's work on it at once. Each bot's **Tasks** tab lists
91
118
  what it is working on.
92
119
  - **What bots say to each other stays out of your way**: in a bot's chat, hand-offs, reports and messages between
93
- bots run together as one quiet group under your conversation with it. **Settings → Chat → Messages between bots**
120
+ bots run together as one quiet group under your conversation with it; those about a task that came from a topic show
121
+ in that topic. **Settings → Chat → Messages between bots**
94
122
  sets how much of them shows, the same on every device: **One line** (the default: who hands what to whom, its
95
123
  stage and time; a click opens the task), **Short** (and the first two lines of the message) or **Full** (the whole
96
124
  message, with the task's card). The bot's replies to you read like the main conversation, without a box.
@@ -109,13 +137,14 @@ for you. You don't open tasks yourself: you ask the dispatcher (or any bot), and
109
137
 
110
138
  The dispatcher works in one of two ways, chosen under **How it works** on its **Profile** tab, for the whole team:
111
139
 
112
- - **Router** (the default): it understands the request, hands the work out and reports back. It does not run
113
- commands, write or edit files, use sub-agents and workflows, draw images or make Office files; it can still read
114
- files, search the web and use its skills. Its **Abilities** section marks what Router mode turns off, and **Turn
140
+ - **Router** (the default): it understands the request, hands the work out and keeps track of it. It does not run
141
+ commands, write or edit files, search the web, use sub-agents and workflows, draw images or make Office files; it can
142
+ still read files and use its skills. Work that needs any of those, when no teammate fits, it hands to itself: it
143
+ then does it in a task of its own, with all its tools, while its own conversation stays free for you. Its **Abilities** section marks what Router mode turns off, and **Turn
115
144
  back on** gives an ability back (even then it may not `sleep`, and a command must finish within 30 seconds or run in
116
- the background). It never waits for a teammate: it ends its turn and is woken when a report arrives, so a new
145
+ the background). It never waits for a teammate: it ends its turn and is woken when something needs it, so a new
117
146
  request is not stuck behind work in progress, and messages that arrive while it is busy are taken together at its
118
- next step. Since it only routes, a light, fast model is enough for it; its **Model** section says so, and Bots does
147
+ next step, each answered in its own topic. Since it only routes, a light, fast model is enough for it; its **Model** section says so, and Bots does
119
148
  not change the model for you. In Router mode every bot that hands work to a teammate opens a task, not only the
120
149
  dispatcher.
121
150
  - **Hands-on**: how 0.7 worked. The dispatcher also does small jobs itself, and a message from one bot's own
@@ -130,10 +159,11 @@ work itself.
130
159
  A task remembers the chat it was asked from: a private chat, a Slack or Discord thread, a Feishu topic or reply chain,
131
160
  a Telegram topic, or the panel.
132
161
 
133
- - **Answers go back where the request came from.** When a task reports, the dispatcher's answer goes to that task's
134
- chat, so two requests from two chats are answered each in its own chat. When one turn of the dispatcher involves
135
- more than one chat (or a chat and the panel), its plain text stays in the panel, marked **Not sent**, and it sends
136
- each answer to its chat with `bot_reply`.
162
+ - **Answers go back where the request came from.** The bot doing a task answers in that task's chat itself, signed
163
+ with its name (`【Forge】`) when the chat app belongs to another bot, so two requests from two chats are answered each
164
+ in its own chat. When one turn of the dispatcher involves more than one chat or topic, it answers each with
165
+ `bot_reply`, and a task it opens there has to say which one it is for. Plain text left over from such a turn stays in
166
+ the panel, marked **Not sent**, unless each chat and topic has its answer already.
137
167
  - **A follow-up in a thread goes straight to the task.** In a Slack or Discord thread, a Feishu topic or reply chain,
138
168
  or a Telegram topic, what you write later in the same thread goes to the tasks started there; the dispatcher only
139
169
  gets a note, which does not wake it. Until its conversation is archived, a done task also takes such a message and
@@ -155,8 +185,9 @@ Every 5 minutes Bots checks the tasks in progress by fixed rules; no model judge
155
185
  - After a step fails with a temporary error (HTTP 408, 429 or 5xx, a dropped or refused connection), the task is told
156
186
  to continue 15 seconds later, then 30, then 60. A fourth such failure in a row, or any other error, and it has
157
187
  **Failed**; the bot that asked is told.
158
- - When every task a bot asked for has settled (done, failed, stopped or stuck) and that bot has said nothing since
159
- (an answer marked **Not sent** doesn't count), it gets one reminder to tell whoever asked. A task that asked for
188
+ - When every task a bot asked for the reports of has settled (done, failed, stopped or stuck) and that bot has said
189
+ nothing since (an answer marked **Not sent** doesn't count), it gets one reminder to tell whoever asked. Tasks that
190
+ answered the person themselves need no reminder. A task that asked for
160
191
  sub-tasks and has ended since gets none.
161
192
  - After Harness restarts, a task that was in the middle of its work gets a note: check what was already done (files,
162
193
  commands, messages sent), don't repeat it, then go on. A task whose conversation is gone starts again from its
@@ -500,6 +531,15 @@ git+https://github.com/kadaliao/dsh-plugins.git#main&path:/plugins/bots
500
531
 
501
532
  Refresh the page once after installing. For AI avatars, install [Images](../images) too.
502
533
 
534
+ ## Upgrading from 0.8
535
+
536
+ - Nothing to do: bots, tasks, memory, schedules, chat apps and pairings stay as they are. The timeline in
537
+ `messages.jsonl` gains fields for topics; older messages stay in the main timeline.
538
+ - A task opened before the upgrade that came from a chat or the panel now answers the person itself when it
539
+ reports, like new ones.
540
+ - The dispatcher in **Router** mode loses web search in its own conversation (it hands searches to a task, its own
541
+ if no teammate fits); **Turn back on** under its **Abilities** gives it back.
542
+
503
543
  ## Upgrading from 0.7
504
544
 
505
545
  - Bots, memory, scheduled tasks, topics, chat apps and pairings stay as they are. 0.8 adds `tickets.json`
@@ -523,7 +563,9 @@ Refresh the page once after installing. For AI avatars, install [Images](../imag
523
563
  - In a group chat a bot hears only messages that @mention it (or reply to it); it does not listen to the
524
564
  rest of the conversation. 0.8 does not change that.
525
565
  - Thread history starts with 0.8: messages from before the upgrade belong to no thread, so reading a chat
526
- thread with `bot_recall` shows only what was said there since.
566
+ thread with `bot_recall` shows only what was said there since. Panel topics start with 0.9.
567
+ - Topics are in the panel only. In a chat app, threads, topics and reply chains play that part where the app has
568
+ them; a private chat stays one conversation, where a new message goes to the dispatcher.
527
569
  - The watchdog goes by time and status only; it does not judge whether a task is heading the right way.
528
570
  - Costs are estimates. Gateways, subscriptions and discounts bill differently; set your own prices.
529
571
  - Only the Codex CLI's login can be borrowed so far; other subscriptions sign in from Harness Settings →
package/README.zh-CN.md CHANGED
@@ -5,6 +5,7 @@
5
5
  把 Grok bot 搬进 DeepSeek Harness:一支能互相协作的 bot 团队。每个 bot 有自己的**头像、人设、模型、技能、能力、应用、定时任务和记忆**,
6
6
  也有自己的会话;bot 之间**私信协作**,其中一个是**调度员**,负责拆解需求、分派、跟进、汇报。
7
7
  分派出去的每件工作都是一个**任务**,bot 在专门的会话里做,所以一个 bot 可以同时做好几件事。
8
+ 和 bot 聊天就像在飞书群里:想到什么随时发,每条消息的回复都收在它自己的**话题**里。
8
9
 
9
10
  它适合两种场景:
10
11
 
@@ -53,6 +54,24 @@ bot 默认用你说话的语言回答;中文提问就是中文协作。
53
54
 
54
55
  上面 Atlas 发给三位队友的每条消息都开了一个任务,三个任务同时进行;Forge 那句话就是它对自己任务的汇报。
55
56
 
57
+ ## 聊天:一件事一个话题
58
+
59
+ 和 bot 聊天就像在飞书群里:想发几条发几条、什么时候发都行,每条消息的回复都收在它自己的话题里。
60
+
61
+ <!-- screenshot: chat with topics -->
62
+
63
+ - **每条消息都是一个话题。** 发出去立刻显示,bot 接手之前标着 **排队中**。对它的回复、它引出的工作的所有动静都收在它下面。
64
+ 消息下方的话题条显示回复数、最新一条、在做这件事的任务(带实时状态),有没看的内容就亮一个点;点开,话题在聊天旁边展开。
65
+ 看不见的话题有了新回复,输入框上方会出现 **N 个话题有新回复**,点一下就过去。
66
+ - **可以回复任何一条消息。** 鼠标移到消息上点 **回复**,你写的话就进那条消息的话题。话题里的任务还没结束(或者一天内刚完成)时,
67
+ 这条话直接交给做任务的 bot;调度员只收到一条提示,不会被唤醒。
68
+ - **调度员一两步之内就回应每一条消息**,因为它不在自己的会话里干活:交给合适的队友,没有合适的就开一个任务交给自己,
69
+ 然后马上空出来。连发两件事会同时推进,不会因为来了新消息就丢掉旧的。接着说某件进行中的事的消息,
70
+ 会并进那个任务,并归入它的话题(**已归入「…」**)。
71
+ - **干活的 bot 直接回复你**,在请求来的那个话题里,给出完整结果。只有调度员把一件事拆给几个 bot、需要汇总结果时,
72
+ 才会让汇报先回到它那里,再由它回复你。
73
+ - 0.9 之前的消息照旧留在主时间线里。
74
+
56
75
  ## 任务:分派出去的工作各有各的会话
57
76
 
58
77
  一个 bot 把工作交给队友,这件工作就成了一个**任务**。队友在专门的会话里做每个任务,所以同一个 bot 可以同时做好几个任务而不串台,
@@ -60,8 +79,10 @@ bot 默认用你说话的语言回答;中文提问就是中文协作。
60
79
 
61
80
  <!-- screenshot: tasks view -->
62
81
 
63
- - **汇报**:做任务的 bot 向发起方汇报,并注明状态:**完成**(附上结果)、**提问**(需要有人回答)、**受阻**(做不下去了),
64
- 或者**进展**——进展只记在任务上,不打扰任何人。完成、提问和受阻会立刻送到发起的 bot,由它转告(调度员会转告给你)。
82
+ - **汇报**:做任务的 bot 汇报时注明状态:**完成**(附上结果)、**提问**(需要有人回答)、**受阻**(做不下去了),
83
+ 或者**进展**——进展只记在任务上,不打扰任何人。完成、提问和受阻原文发给你,发在请求来的那个话题或聊天里;
84
+ 发起的 bot 只收到一条提示,不会被唤醒。发起的 bot 要求汇报先回给它时(为了汇总几份结果),汇报立刻送到它那里,由它转告;
85
+ 子任务的汇报、以及不来自任何聊天的任务(比如定时任务开出的)的汇报也是这样。
65
86
  任务结束后才发来的汇报,或者所属任务已经结束的子任务发来的汇报,只记在任务上。
66
87
  做任务的 bot 写的其他内容都留在任务自己的会话里。
67
88
  - **子任务**:做任务的 bot 可以找队友帮忙,这会开一个子任务,子任务的汇报回到发起它的那个任务里。
@@ -75,7 +96,8 @@ bot 默认用你说话的语言回答;中文提问就是中文协作。
75
96
  还有 **打开会话**、**重试**、**标为完成**、**停止** 几个按钮,以及一个给任务发消息的输入框(比如回答它的提问);
76
97
  **标为完成** 和 **停止** 都会结束任务,并让 bot 立刻停下手上的活。
77
98
  每个 bot 的 **任务** 标签页列出它手上的工作。
78
- - **bot 之间的消息不抢戏**:在 bot 的聊天里,分派、汇报和 bot 之间的消息连在一起,收成你和它对话下面安静的一组。
99
+ - **bot 之间的消息不抢戏**:在 bot 的聊天里,分派、汇报和 bot 之间的消息连在一起,收成你和它对话下面安静的一组;
100
+ 来自某个话题的任务,相关的这些消息显示在那个话题里。
79
101
  **设置 → 聊天 → Bot 之间的消息** 决定显示多少,每台设备都一样:**一行**(默认:谁把什么交给谁、进度和时间,点一下打开任务)、
80
102
  **简短**(再加消息的前两行)、**完整**(整条消息,委派附任务卡片)。bot 回复你的话和主对话一样排版,不再套框。
81
103
  - **实时状态**:bot 干活时,聊天里会写出它此刻在做什么、做了多久("运行 `npm test` · 42 秒"、"思考中 · 8 秒",再加第几步和总用时);
@@ -89,12 +111,13 @@ bot 默认用你说话的语言回答;中文提问就是中文协作。
89
111
 
90
112
  调度员有两种工作方式,在它的 **资料** 页 **调度方式** 里选,对整个团队生效:
91
113
 
92
- - **只调度**(默认):听懂需求、把工作分派出去、汇报结果,自己不动手。这时它不执行命令、不写不改文件、不用子代理和工作流、
93
- 不作图、不做 Office 文档,但仍能读文件、联网搜索、使用技能。**能力** 一栏会标出只调度时关掉的项,
114
+ - **只调度**(默认):听懂需求、把工作分派出去、跟进进度,自己不动手。这时它不执行命令、不写不改文件、不联网搜索、
115
+ 不用子代理和工作流、不作图、不做 Office 文档,但仍能读文件、使用技能。需要这些能力的事,没有合适的队友时,
116
+ 它就交给自己:在一个单独的任务里用全部工具去做,它自己的会话始终空着等你。**能力** 一栏会标出只调度时关掉的项,
94
117
  点 **重新打开** 可以逐项放回(即使放回了命令,它也不能 `sleep`,命令要么在 30 秒内跑完,要么放到后台跑)。
95
- 它从不等队友:分派完就结束这一轮,汇报到了会自动唤醒它,所以新请求不会被手上的活堵住;
96
- 它忙的时候进来的消息,会在下一步一起处理。既然只管调度,用轻量、快速的模型就够了,**模型** 一栏也会这样提示;
97
- Bots 不会替你换模型。只调度时,任何 bot 把工作交给队友都会开任务,不只调度员。
118
+ 它从不等队友:分派完就结束这一轮,有事需要它时会自动唤醒,所以新请求不会被手上的活堵住;
119
+ 它忙的时候进来的消息,会在下一步一起处理,各自在自己的话题里得到回应。既然只管调度,用轻量、快速的模型就够了,
120
+ **模型** 一栏也会这样提示;Bots 不会替你换模型。只调度时,任何 bot 把工作交给队友都会开任务,不只调度员。
98
121
  - **亲自动手**:0.7 的做法。调度员小事自己顺手做掉;bot 在自己的会话里给队友发消息,消息直接进队友自己的会话,不开任务。
99
122
  已经开出的任务照常进行、照常汇报。
100
123
 
@@ -104,9 +127,9 @@ bot 默认用你说话的语言回答;中文提问就是中文协作。
104
127
 
105
128
  任务会记住它来自哪个聊天:私聊、Slack 讨论串、Discord 子区或回复链、飞书话题或回复链、Telegram 话题,或者面板。
106
129
 
107
- - **结果回到请求来的地方。** 任务一汇报,调度员的答复就发到这个任务来自的聊天,所以从两个聊天各派一件事,结果各回各的聊天。
108
- 如果调度员的某一轮同时涉及好几个聊天(或者一个聊天加面板),它的普通回复只留在面板里,标着 **未发出**;
109
- 它会用 `bot_reply` 把每条答复发到各自的聊天。
130
+ - **结果回到请求来的地方。** 做任务的 bot 自己把结果发到这个任务来自的聊天;聊天应用属于别的 bot 时,开头署上它的名字(`【Forge】`),
131
+ 所以从两个聊天各派一件事,结果各回各的聊天。如果调度员的某一轮同时涉及好几个聊天或话题,它会用 `bot_reply` 分别回复,
132
+ 这一轮开的任务也要说明是给哪一个的。这一轮剩下的普通回复只留在面板里,标着 **未发出**;每个聊天和话题都已经回复过的话就不显示。
110
133
  - **讨论串里的跟进直接交给任务。** 在 Slack 讨论串、Discord 子区或回复链、飞书话题或回复链、Telegram 话题里,
111
134
  你之后在同一个讨论串里说的话,直接交给从这里开出的任务;调度员只收到一条提示,不会被唤醒。
112
135
  在会话归档之前,已完成的任务收到这样的消息也会接着做;已停止的任务不会。
@@ -124,8 +147,8 @@ Bots 每 5 分钟按固定规则检查一遍进行中的任务,不靠模型判
124
147
  再过 20 分钟还没汇报,就标为 **卡住了**,并通知发起的 bot。
125
148
  - 某一步遇到临时错误(HTTP 408、429 或 5xx,连接断开或被拒),15 秒后让任务继续,再失败就等 30 秒、再等 60 秒;
126
149
  连续第四次失败,或者遇到其他错误,任务就 **失败**,并通知发起的 bot。
127
- - 一个 bot 发起的任务全都有了结果(已完成、失败、已停止或卡住了),它却一直没开口(标着 **未发出** 的回复不算),
128
- 就提醒它一次,让它告诉请求方。发起方如果是一个已经结束的任务,就不再提醒。
150
+ - 一个 bot 要求回报的任务全都有了结果(已完成、失败、已停止或卡住了),它却一直没开口(标着 **未发出** 的回复不算),
151
+ 就提醒它一次,让它告诉请求方。发起方如果是一个已经结束的任务,就不再提醒;直接回复了你的任务也不需要提醒。
129
152
  - Harness 重启后,做到一半的任务会收到一条说明:先检查已经做过的事(文件、命令、发出去的消息),不要重复,再接着做。
130
153
  会话已经不在的任务,带着同一条说明从原来的交代重新开始;交代没送进会话的任务,会重新收到交代。
131
154
  排队的任务和等名额的消息,有空位时再开始、再送达;同一条消息不会投递两次。
@@ -404,6 +427,13 @@ git+https://github.com/kadaliao/dsh-plugins.git#main&path:/plugins/bots
404
427
 
405
428
  安装后刷新一次页面。想要 AI 头像,再装上[图片插件](../images)。
406
429
 
430
+ ## 从 0.8 升级
431
+
432
+ - 不用做什么:bot、任务、记忆、定时任务、聊天应用和配对都原样保留。时间线 `messages.jsonl` 多了话题相关的字段,
433
+ 以前的消息照旧留在主时间线。
434
+ - 升级前开出、来自某个聊天或面板的任务,之后汇报时也直接回复你,和新任务一样。
435
+ - 「只调度」的调度员在自己的会话里不再联网搜索(搜索交给任务,没有合适的队友就交给自己);在它的 **能力** 一栏点 **重新打开** 可以放回。
436
+
407
437
  ## 从 0.7 升级
408
438
 
409
439
  - bot、记忆、定时任务、话题、聊天应用和配对都原样保留;0.8 新增 `tickets.json` 存放任务。
@@ -421,7 +451,8 @@ git+https://github.com/kadaliao/dsh-plugins.git#main&path:/plugins/bots
421
451
  - 仅 Web 界面。
422
452
  - 暂无 bot 群聊,bot 之间通过任务和私信协作。
423
453
  - 在群聊里,bot 只听 @ 它(或回复它)的消息,不旁听群里的其他对话;0.8 没有改变这一点。
424
- - 讨论串的历史从 0.8 开始:升级前的消息不属于任何讨论串,所以用 `bot_recall` 读某个聊天讨论串时,只能看到升级以后的内容。
454
+ - 讨论串的历史从 0.8 开始:升级前的消息不属于任何讨论串,所以用 `bot_recall` 读某个聊天讨论串时,只能看到升级以后的内容。面板话题从 0.9 开始。
455
+ - 话题只在面板里有。聊天应用里,由应用自己的讨论串、话题和回复链起这个作用;私聊仍是一整段对话,新消息交给调度员。
425
456
  - 看门狗只看时间和状态,不判断任务的方向对不对。
426
457
  - 费用是估算,网关、订阅和折扣的实际计费会不同;可以填自己的价格。
427
458
  - 目前只能借用 Codex CLI 的登录;其他订阅请在 Harness 支持的情况下,到 设置 → 模型 里登录。