feat(messaging): compact conversation rows and relabel External connectors (#1113)

CHANGELOG.md

@@ -41,6 +41,8 @@ WeChat original voice can be prepared for playback, native quotes and retained l
 
 ### Changed
 
+- The External connectors entry no longer offers **Add connector** or **Authorize a conversation** while syncing is redesigned around connecting any app's conversation; it lists existing syncs and keeps **Save a send target (advanced)** for apps that can't post to conversations directly. A conversation row shows where it is synced, and its actions (**Mute**, **Rules**, **Block**) sit on their own row with a tooltip each. The Lark setup guide shrinks to three steps (connect the app, bind it, send a test message), and the Lark, Slack and WeChat guides are rewritten around **Bind app** ([#1113](https://github.com/BotHarness/BotHarness/issues/1113), [Lark guide](docs/lark-connection.md), [External connectors guide](docs/channel-sidebar/external-connectors.md)).
+
 - A Bot's model and wake policy moved from its Profile to two new entries in the DM's Channel sidebar, **Model** and **Wake policy**. Each shows its current value in the collapsed header, lists one row per setting, and opens a dialog to edit; both can be reordered or hidden like other entries. The Model dialog now lists a **Main model** and a **Task model**, each picked from a searchable list with its reasoning effort; models can be set without any preset, a saved preset can still fill the dialog, and **Save as preset** keeps the current settings for reuse. **Attention policy** is now called **Wake policy** ([#1084](https://github.com/BotHarness/BotHarness/issues/1084), [Model guide](docs/channel-sidebar/model.md), [Wake policy guide](docs/channel-sidebar/wake-policy.md)).
 - A Bot's IM settings moved from its Profile to two new entries in the DM's Channel sidebar. **External identities** lists each bound identity as a row with its availability and an on/off switch, plus **Bind identity**, the Lark setup guide, IM administrator pairing and Lark approval notifications. **External connectors** lists one row per connector with its state and an on/off switch, plus **Add connector** and conversation authorization. Clicking a row opens a dialog to edit it, with the same options as before. **Channel connectors** are now called **External connectors**. The SOUL.md and MEMORY.md standing limits became one row under **Memory files** that opens a dialog. The moved forms now use the shared dropdowns, inputs and checkboxes ([#1085](https://github.com/BotHarness/BotHarness/issues/1085), [External identities guide](docs/channel-sidebar/external-identities.md), [External connectors guide](docs/channel-sidebar/external-connectors.md)).
 

CHANGELOG.zh.md

@@ -41,6 +41,8 @@
 
 ### Changed
 
+- 「外部连接器」不再提供 **添加外部连接器** 和 **授权外部会话**（同步正在围绕“接入任意应用的会话”重新设计），只列出已有的同步，并保留 **保存发送目标（高级）** 给不能直接发往会话的应用。会话行会显示已同步到哪里，操作按钮（**静音**、**规则**、**屏蔽**）单独一行并各带说明提示。Lark 配置引导缩减为三步（连接应用、绑定、发消息验证），Lark、Slack、微信连接教程围绕 **绑定应用** 重写（[#1113](https://github.com/BotHarness/BotHarness/issues/1113)、[Lark 教程](docs/lark-connection.md)、[外部连接器说明](docs/channel-sidebar/external-connectors.md)）。
+
 - Bot 的模型和唤醒策略从 Profile 移到私聊 Channel sidebar 的两个新分区 **模型** 与 **唤醒策略**：收起时标题右侧显示当前值，每项设置一行，点击打开弹窗修改，并可像其他分区一样排序或隐藏。模型弹窗改为 **主模型** 与 **执行模型** 两项，可输入过滤选择模型并选思考强度；不建预设也能直接保存，已有预设仍可用来填入，**另存为预设** 可把当前设置留作复用。**提醒策略** 改名为 **唤醒策略**（[#1084](https://github.com/BotHarness/BotHarness/issues/1084)，[模型教程](docs/channel-sidebar/model.md)，[唤醒策略教程](docs/channel-sidebar/wake-policy.md)）。
 - Bot 的 IM 设置从 Profile 移到私聊 Channel sidebar 的两个新分区。**外部身份** 每个已绑定身份一行，显示可用状态和启用开关，下面是 **绑定身份**、Lark 配置引导、IM 管理员配对和 Lark 审批通知。**外部连接器** 每个连接器一行，显示状态和启用开关，下面是 **添加外部连接器** 和外部会话授权。点击一行打开弹窗修改，选项与原来一致。**频道连接器** 改名为 **外部连接器**。SOUL.md 和 MEMORY.md 的常驻记忆上限变成 **记忆文件** 下的一行，点击打开弹窗。迁移后的表单统一使用共享的下拉框、输入框和复选框（[#1085](https://github.com/BotHarness/BotHarness/issues/1085)，[外部身份教程](docs/channel-sidebar/external-identities.md)，[外部连接器教程](docs/channel-sidebar/external-connectors.md)）。
 

CONTEXT.md

@@ -435,9 +435,25 @@ The membership state of a Channel that belongs to no Channel section. Ungrouped
 _Avoid_: default folder, inbox, fixed bottom bucket
 
 **Bridge**:
-A configured connection from an external source, such as an IM conversation or later a webhook, to an explicit Channel or PersonaBot Inbox target; it carries inbound delivery and exposes outbound capabilities without becoming the Actor. The UI calls it an External connector (外部连接器); the PersonaBot's own external identity is a separate thing.
+A configured connection from an external source, such as an IM conversation or later a webhook, to an explicit Channel or PersonaBot Inbox target; it carries inbound delivery and exposes outbound capabilities without becoming the Actor. For an IM conversation it is listed under External connectors (外部连接器); the PersonaBot's own external identity is a separate thing.
 _Avoid_: integration, adapter, connector (bare)
 
+**App**:
+The UI name (应用) for one authenticated Provider account, such as a Lark app, a Slack app, a Discord bot or a paired WeChat Bot. An App is bound to at most one PersonaBot; a PersonaBot may bind several Apps, including several of one platform. Settings → IM apps lists every App and the Bot that uses it.
+_Avoid_: IM account (in UI copy), connector, integration
+
+**Default traffic**:
+The messages an enabled external identity Binding admits without any per-conversation consent: direct messages to the App and group messages the qualified Provider reports as mentioning it. Where the App can be reached is decided on the platform; ordinary group text, followed threads and Channel syncs stay explicit.
+_Avoid_: all messages, authorized traffic, open intake
+
+**Conversation entry**:
+The per-conversation anchor of one Binding's traffic, stored as a Messaging Grant. An _implicit_ entry is recorded on the first admitted default-traffic message or first proactive post; an _explicit_ entry comes from a saved send target. A _held_ entry waits for the Human (Ask me first, or a bound was reached) and keeps only metadata. A _blocked_ conversation has a durable block keyed by Bot, App fingerprint, kind and ID that refuses it until **Allow again**. Muting an entry keeps admission but never wakes the Bot.
+_Avoid_: authorization (for implicit entries), delivery target, subscription
+
+**New-conversation mode**:
+A Binding's choice for a conversation that has no entry yet: `auto` (Admit automatically, the default) records an implicit entry and admits the message; `ask` (Ask me first) holds it. Auto-creation is bounded at 20 new entries per hour and 500 active entries per Binding.
+_Avoid_: whitelist, approval mode, auto-reply
+
 **Bot Inbox**:
 The PersonaBot-level view of Source Events admitted for its attention, whether or not an event belongs to a Channel. It is not a second content store: reading is an explicit act, and ignoring is allowed.
 _Avoid_: queue, mailbox, backlog

CONTEXT.zh.md

@@ -433,9 +433,25 @@ Channel 不属于任何 Channel section 时的 membership state。未分组 Chan
 _避免使用_：default folder、inbox、fixed bottom bucket
 
 **Bridge**：
-从 external source（如 IM 会话，未来还有 webhook）到某个显式 Channel 或 PersonaBot Inbox target 的已配置连接；它承载 inbound delivery 并暴露 outbound capability，但不会成为 Actor。UI 上称为「外部连接器 / External connector」；PersonaBot 自己的外部身份是另一回事。
+从 external source（如 IM 会话，未来还有 webhook）到某个显式 Channel 或 PersonaBot Inbox target 的已配置连接；它承载 inbound delivery 并暴露 outbound capability，但不会成为 Actor。对于 IM 会话，它列在「外部连接器 / External connector」中；PersonaBot 自己的外部身份是另一回事。
 _避免使用_：integration、adapter、裸用 connector
 
+**App**：
+UI 上叫「应用」：一个已认证 Provider account 的名称，例如 Lark 应用、Slack 应用、Discord bot 或已配对的微信 Bot。一个应用最多绑定到一个 PersonaBot；一个 PersonaBot 可以绑定多个应用，包括同一平台的多个应用。「设置 → IM 应用」列出每个应用及使用它的 Bot。
+_避免使用_：UI 文案里的「IM账号」、connector、integration
+
+**Default traffic**：
+启用的外部身份 Binding 无需逐会话同意即可接收的消息：发给应用的私聊，以及合格 Provider 报告为 @ 该应用的群消息。谁能触达应用由平台决定；群里的普通文字、跟进的话题和 Channel 同步仍需显式开启。
+_避免使用_：全部消息、已授权流量、开放收件
+
+**Conversation entry**：
+一个 Binding 的单个会话流量所挂靠的锚点，存为 Messaging Grant。_隐式_ 条目在第一条被接收的默认流量消息或第一次主动发送时记录；_显式_ 条目来自保存的发送目标。_等待处理_ 的条目等 Human 决定（先问我，或达到上限），只保留元数据。_已屏蔽_ 的会话有一个按 Bot、应用指纹、会话类型和 ID 记录的持久屏蔽，直到 **再次允许** 前都会被拒绝。静音的条目照常接收，但不会唤醒 Bot。
+_避免使用_：授权（指隐式条目时）、投递目标、订阅
+
+**New-conversation mode**：
+Binding 对尚无条目的会话的处理方式：`auto`（自动接收，默认）记录隐式条目并接收消息；`ask`（先问我）把它放到等待处理。自动创建有上限：每个 Binding 每小时 20 个新条目、同时 500 个活跃条目。
+_避免使用_：白名单、审批模式、自动回复
+
 **Bot Inbox**：
 PersonaBot 层级的 view，包含被 admit 供其 attention 的 Source Event，无论 event 是否属于某个 Channel。它不是第二个 content store：读取是一项显式行为，也允许 ignore。
 _避免使用_：queue、mailbox、backlog

docs/architecture/botharness-architecture.en.md

@@ -219,10 +219,14 @@ sequenceDiagram
   M->>DB: append attempt and outcome facts
 ```
 
-A Source Event is the sole authority for content; Channels and Inboxes hold relationships only. A Reply follows the trusted Reply Route so the Host selects the source provider. A proactive post is a Service Action and requires both Provider Capability and a Human Service Grant. A SQLite transaction covers local facts only. External effects use an Outbox Intent, a stable idempotency identity, and bounded reconciliation without claiming exactly-once delivery. An outcome that cannot be proven becomes `unknown-outcome` for Human resolution.
+A Source Event is the sole authority for content; Channels and Inboxes hold relationships only. A Reply follows the trusted Reply Route so the Host selects the source provider. A proactive post is a Service Action and requires Provider Capability plus either the Bot's conversation entry for a conversation the App can reach or, as a fallback, a Human saved-target Service Grant. A SQLite transaction covers local facts only. External effects use an Outbox Intent, a stable idempotency identity, and bounded reconciliation without claiming exactly-once delivery. An outcome that cannot be proven becomes `unknown-outcome` for Human resolution.
 
 Wake Policy decides when an Orchestrator observes new attention: at the safe boundary after the current step, after the current turn, or by starting a new turn while idle. Ordinary external messages do not interrupt a running model/tool step. Only a DSH-supported and policy-authorized control path may steer execution. Ready attention is consumed by turn, not by event: while a turn runs, arrivals mark a ready set, and one harvest turn consumes it when the turn ends or the Bot is idle; only direct mentions and DMs steer into the running turn (ADR-0077). Wake handling follows the Source class rather than the platform: external providers normalize into Source Events at the Bridge, and the runtime branches only on Source class and admission reason (ADR-0075).
 
+### Bound apps admit default traffic (ADR-0142)
+
+Reception is keyed by the external identity Binding, not by a saved target. An enabled Binding admits its App's default traffic, direct messages and qualified group mentions, into the bound Bot's Inbox. The first admitted message from a conversation commits an implicit conversation entry (a `messaging_grants` row with `origin: 'implicit'`) in the same transaction as its Source Event and Inbox Admission, so Outbox fences, group policies, Channel Bridge routes and revocation keep one anchor. The Binding's new-conversation mode (`auto` or `ask`) and its bounds (20 new entries per hour, 500 active, 200 held) decide whether a new conversation is admitted or held. A block is durable per Bot, App fingerprint, conversation kind and ID; mute moves only `preferenceRevision`, leaving the entry's Outbox revision untouched. A Bot may bind several Apps of one platform; each App still belongs to one Bot. In the Client the App's conversation list manages entries and shows their existing Channel Bridge routes (creating a sync is being redesigned around External connectors), and the External connectors entry lists those routes and keeps saved-target Service Grants as the advanced fallback for Providers that can't list or address conversations.
+
 ### Local Human names — target design (ADR-0103)
 
 The local Human has one optional default display name within the DSH Profile, edited in BotHarness plugin settings and falling back to `Human`. Every Human-participating DM or Group Channel may override it with **Human Channel nickname**, edited through “My nickname” in the Channel header menu. Clearing an override restores inheritance; changing the default affects only Channels without an override. These names support roleplay with different partners without adding saved roleplay backgrounds, another Human account, or another Human Inbox.

docs/architecture/botharness-architecture.md

@@ -227,10 +227,14 @@ sequenceDiagram
   M->>DB: append attempt and outcome facts
 ```
 
-Source Event 是内容唯一权威；Channel 和 Inbox 都只保存关系。Reply 使用可信 Reply Route 自动选择来源 provider；主动发布属于 Service Action，必须同时满足 Provider Capability 与 Human Service Grant。SQLite 事务只覆盖本地事实；外部副作用使用 Outbox Intent、幂等标识和有界 reconciliation，不宣称 exactly-once。不可证明的结果进入 `unknown-outcome`，由 Human 处理。
+Source Event 是内容唯一权威；Channel 和 Inbox 都只保存关系。Reply 使用可信 Reply Route 自动选择来源 provider；主动发布属于 Service Action，需要 Provider Capability，以及该 Bot 对应用可触达会话的会话条目，或作为后备的 Human 保存目标 Service Grant。SQLite 事务只覆盖本地事实；外部副作用使用 Outbox Intent、幂等标识和有界 reconciliation，不宣称 exactly-once。不可证明的结果进入 `unknown-outcome`，由 Human 处理。
 
 Wake Policy 决定何时让 Orchestrator 看见新 attention：当前 step 完成后的安全边界、当前 turn 结束后，或 idle 时启动新 turn。普通外部消息不打断正在执行的 model/tool step；只有 DSH 明确支持且策略授权的控制路径才能 steer。就绪的 attention 按回合收割：忙碌期间新到的事件只把就绪集合置脏，当前回合结束（即空闲）时由一次 harvest turn 消费全部就绪项；主动 steer 只用于直接 @ 与 DM（ADR-0077）。
 
+### 绑定的应用接收默认流量（ADR-0142）
+
+接收以外部身份 Binding 为键，而不是保存的目标。启用的 Binding 把应用的默认流量（私聊和合格的群 @）接收进绑定 Bot 的 Inbox。会话的第一条被接收的消息，会与其 Source Event、Inbox Admission 在同一事务中提交一个隐式会话条目（`messaging_grants` 中 `origin: 'implicit'` 的行），使 Outbox fence、群策略、Channel Bridge 路由和撤销仍挂在同一个锚点上。Binding 的新会话模式（`auto` 或 `ask`）及其上限（每小时 20 个新条目、500 个活跃、200 个等待）决定新会话被接收还是等待处理。屏蔽按 Bot、应用指纹、会话类型和 ID 持久保存；静音只推进 `preferenceRevision`，不改动条目的 Outbox revision。一个 Bot 可以绑定同一平台的多个应用，每个应用仍只属于一个 Bot。Client 中，应用的会话列表管理这些条目并显示已有的 Channel Bridge 路由（新建同步正在围绕外部连接器重新设计），「外部连接器」列出这些路由并保留保存目标的 Service Grant，作为 Provider 无法列出或直接发往会话时的高级后备。
+
 [#824](https://github.com/BotHarness/BotHarness/issues/824) 在 PersonaBot Profile 增加可恢复的配置引导，本地打包的 Driver.js 定位既有原生 IM 设置、身份与群授权控件。进度只投影当前兼容 Provider、账号、Binding／Grant、已收测试 Source Event 和沿原来源回复的 Outbox 回执；可选的认证自身回传仍单独呈现。平台接受、外部可见投递和已读回执明确区分。不建立新手引导数据库、额外凭据存储或消息权威；打开和关闭引导不产生授权或发送。
 
 ### 首条外部出站路径（ADR-0101）

docs/assets/pr/1113-connectors-relabel/README.md

@@ -0,0 +1,24 @@
+# #1113 External connectors relabel
+
+Captured from a real DSH 0.2.0-rc.1 instance with the QA fake IM provider and the real OpenCode Go model.
+
+## Conversation rows: actions on their own row, a tooltip on each
+
+|     | Light                      | Dark                      |
+| --- | -------------------------- | ------------------------- |
+| zh  | ![](rows-v4-zh-light.webp) | ![](rows-v4-zh-dark.webp) |
+| en  | ![](rows-v4-en-light.webp) | ![](rows-v4-en-dark.webp) |
+
+## External connectors: syncs plus the advanced send target
+
+|     | Light                    | Dark                    |
+| --- | ------------------------ | ----------------------- |
+| zh  | ![](conns-zh-light.webp) | ![](conns-zh-dark.webp) |
+| en  | ![](conns-en-light.webp) | ![](conns-en-dark.webp) |
+
+## Lark setup guide: three steps
+
+|     | Light                       | Dark                       |
+| --- | --------------------------- | -------------------------- |
+| zh  | ![](guide-v2-zh-light.webp) | ![](guide-v2-zh-dark.webp) |
+| en  | ![](guide-v2-en-light.webp) | ![](guide-v2-en-dark.webp) |