npm - @qwen-code/qwen-code - Versions diffs - 0.23.2-preview.0 → 0.23.2 - Mend

@qwen-code/qwen-code 0.23.2-preview.0 → 0.23.2

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.

Files changed (282) hide show

package/bundled/qc-helper/docs/configuration/auth.md CHANGED Viewed

@@ -2,7 +2,7 @@
 Qwen Code's first-run `/auth` menu has three top-level options. Pick the one that matches how you want to run the CLI:
-- **Alibaba ModelStudio**: official recommended setup. Opens a sub-menu with **Coding Plan** (for individual developers · weekly quota included), **Token Plan** (for teams and companies · usage-based billing with a dedicated endpoint), or **Standard API Key** (connect with an existing ModelStudio API key).
+- **Alibaba ModelStudio**: official recommended setup. Opens a sub-menu with **Coding Plan** (for individual developers · weekly quota included), **Token Plan** (for teams and companies · usage-based billing with a dedicated endpoint), or **Standard API Key** (connect with an existing ModelStudio API key). Token Plan and Standard API Key also enable the built-in [`web_search` tool](../../developers/tools/web-search.md) with no extra setup.
 - **Third-party Providers**: choose a built-in provider and connect with an API key (DeepSeek, Grok, MiniMax, Z.AI, Kimi, Idealab, ModelScope, OpenRouter, Requesty).
 - **Custom Provider**: manually connect a local server, proxy, or unsupported provider — supports OpenAI, Anthropic, Gemini, and other compatible endpoints.

package/bundled/qc-helper/docs/configuration/model-providers.md CHANGED Viewed

@@ -666,7 +666,7 @@ The merge strategy for `modelProviders` itself is REPLACE: the entire `modelProv
 ## Reasoning / thinking configuration
-The optional `reasoning` field under `generationConfig` controls how aggressively the model reasons before responding. The Anthropic and Gemini converters always honor it. The OpenAI-compatible pipeline honors it **unless** `generationConfig.samplingParams` is set — see the "Interaction with `samplingParams`" caveat below.
+The optional `reasoning` field under `generationConfig` controls how aggressively the model reasons before responding. The Anthropic and Gemini converters always honor it. The OpenAI-compatible pipeline honors it **unless** `generationConfig.samplingParams` is set. Known GPT-5 models and GPT-6 Astra are an exception: unrelated sampling keys do not suppress configured effort. See "Interaction with `samplingParams`" below.
 ```jsonc
 {
@@ -693,33 +693,37 @@ The optional `reasoning` field under `generationConfig` controls how aggressivel
 ### Per-provider behavior
-| Protocol / provider                           | Wire shape                                                           | Notes                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                       |
-| --------------------------------------------- | -------------------------------------------------------------------- | ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
-| **OpenAI / DashScope** (`qwen3.8-max` family) | Flat `reasoning_effort: <effort>` body parameter                     | The `/effort` tiers are passed through for any model id starting with `qwen3.8-max` (including dated snapshots and `-latest` aliases); DashScope applies any model-specific mapping. This family's ladder stops at `xhigh`, so a configured `max` is clamped to `xhigh` (logged once) rather than sent and rejected. An explicit `reasoning_effort` in `samplingParams` or `extra_body` is a verbatim override and is not clamped. When `reasoning_effort` and `thinking_budget` conflict, the normal `extra_body` > `samplingParams` > `reasoning` precedence keeps only the higher-priority field; an explicit same-layer pair keeps `reasoning_effort`, matching the provider's behavior before cross-layer resolution. If a static field wins, `/effort` reports that field instead of implying the requested tier is effective. When an effort tier wins, a conflicting `enable_thinking` is also dropped. An explicit `enable_thinking: false` in `extra_body` is honoured rather than dropped: it overrides the configured tier as `reasoning_effort: 'none'`, one of the few places `extra_body` does not win verbatim. Other Qwen models continue to map a selected effort to `enable_thinking: true`; a `reasoning_effort` override passes through there unless it conflicts with a `thinking_budget` (a pair DashScope rejects), in which case the inert `reasoning_effort` is dropped and both `enable_thinking` and `thinking_budget` survive. |
-| **OpenAI / DeepSeek** (`api.deepseek.com`)    | Flat `reasoning_effort: <effort>` body parameter                     | When `reasoning.effort` is set in the nested config shape, it's rewritten to flat `reasoning_effort` and `'low'`/`'medium'` are normalized to `'high'`, `'xhigh'` to `'max'` — mirroring DeepSeek's [server-side back-compat](https://api-docs.deepseek.com/zh-cn/api/create-chat-completion). Top-level `samplingParams.reasoning_effort` or `extra_body.reasoning_effort` overrides skip this normalization and ship verbatim. `max` is accepted only on a real DeepSeek hostname; a `deepseek`-named model on another host keeps the generic `xhigh` ceiling, matching the hostname gate on the reshape itself.                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                          |
-| **OpenAI / Z.ai** (`z.ai`, `bigmodel.cn`)     | Flat `reasoning_effort: <effort>` body parameter                     | GLM-5.2+ on a Z.ai host takes the full ladder, `max` included, and the nested `reasoning.effort` is rewritten to the flat field. Older GLM ids, and a `glm-*` model reached on any other host, keep the generic `xhigh` ceiling: the model name alone says nothing about what that endpoint accepts.                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                        |
-| **OpenAI** (other compatible servers)         | `reasoning: { effort, ... }` passed through                          | A configured `max` is clamped to `xhigh` (logged once), since `max` is a vendor extension rather than part of the generic OpenAI ladder. Set via `samplingParams` (e.g. `samplingParams.reasoning_effort` for GPT-5/o-series) when the provider expects a different shape; an explicit `samplingParams` / `extra_body` value is not clamped.                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                |
-| **Anthropic** (real `api.anthropic.com`)      | `output_config: { effort }` plus the `effort-2025-11-24` beta header | Real Anthropic accepts `'low'`/`'medium'`/`'high'` only. `'max'` is **clamped to `'high'`** with a `debugLogger.warn` line (once per generator); if you want max effort, switch the baseURL to a DeepSeek-compatible endpoint that supports it.                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                             |
-| **Anthropic** (`api.deepseek.com/anthropic`)  | Same `output_config: { effort }` + beta header                       | `'max'` is passed through unchanged.                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                        |
-| **Gemini** (`@google/genai`)                  | `thinkingConfig: { includeThoughts: true, thinkingLevel }`           | `'low'` → `LOW`, `'high'`/`'max'` → `HIGH`, others → `THINKING_LEVEL_UNSPECIFIED` (Gemini has no `MAX` tier).                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                               |
+| Protocol / provider                           | Wire shape                                                                           | Notes                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                       |
+| --------------------------------------------- | ------------------------------------------------------------------------------------ | ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
+| **OpenAI / DashScope** (`qwen3.8-max` family) | Flat `reasoning_effort: <effort>` body parameter                                     | The `/effort` tiers are passed through for any model id starting with `qwen3.8-max` (including dated snapshots and `-latest` aliases); DashScope applies any model-specific mapping. This family's ladder stops at `xhigh`, so a configured `max` is clamped to `xhigh` (logged once) rather than sent and rejected. An explicit `reasoning_effort` in `samplingParams` or `extra_body` is a verbatim override and is not clamped. When `reasoning_effort` and `thinking_budget` conflict, the normal `extra_body` > `samplingParams` > `reasoning` precedence keeps only the higher-priority field; an explicit same-layer pair keeps `reasoning_effort`, matching the provider's behavior before cross-layer resolution. If a static field wins, `/effort` reports that field instead of implying the requested tier is effective. When an effort tier wins, a conflicting `enable_thinking` is also dropped. An explicit `enable_thinking: false` in `extra_body` is honoured rather than dropped: it overrides the configured tier as `reasoning_effort: 'none'`, one of the few places `extra_body` does not win verbatim. Other Qwen models continue to map a selected effort to `enable_thinking: true`; a `reasoning_effort` override passes through there unless it conflicts with a `thinking_budget` (a pair DashScope rejects), in which case the inert `reasoning_effort` is dropped and both `enable_thinking` and `thinking_budget` survive. |
+| **OpenAI / DeepSeek** (`api.deepseek.com`)    | Flat `reasoning_effort: <effort>` body parameter                                     | When `reasoning.effort` is set in the nested config shape, it's rewritten to flat `reasoning_effort` and `'low'`/`'medium'` are normalized to `'high'`, `'xhigh'` to `'max'` — mirroring DeepSeek's [server-side back-compat](https://api-docs.deepseek.com/zh-cn/api/create-chat-completion). Top-level `samplingParams.reasoning_effort` or `extra_body.reasoning_effort` overrides skip this normalization and ship verbatim. `max` is accepted only on a real DeepSeek hostname; a `deepseek`-named model on another host keeps the generic `xhigh` ceiling, matching the hostname gate on the reshape itself.                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                          |
+| **OpenAI / Z.ai** (`z.ai`, `bigmodel.cn`)     | Flat `reasoning_effort: <effort>` body parameter                                     | GLM-5.2+ on a Z.ai host takes the full ladder, `max` included, and the nested `reasoning.effort` is rewritten to the flat field. Older GLM ids, and a `glm-*` model reached on any other host, keep the generic `xhigh` ceiling: the model name alone says nothing about what that endpoint accepts.                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                        |
+| **OpenAI** (other compatible servers)         | Known GPT-5 / GPT-6 Astra: flat `reasoning_effort`; other models: nested `reasoning` | GPT effort is clamped in both directions to the supported subset for the known model. GPT-5.6 and GPT-6 Astra allow `max`, while earlier models have lower ceilings. Tiers below the model floor are raised: GPT-5 Pro accepts only `high`; GPT-5.2 Pro, GPT-5.4 Pro and GPT-5.5 Pro raise `low` to `medium`. OpenRouter retains nested `reasoning`. Unknown model names keep the generic `xhigh` ceiling and nested shape. Explicit reasoning values in `samplingParams` / `extra_body` bypass the configured-tier clamp.                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                  |
+| **Anthropic** (real `api.anthropic.com`)      | `output_config: { effort }` plus the `effort-2025-11-24` beta header                 | Real Anthropic accepts `'low'`/`'medium'`/`'high'` only. `'max'` is **clamped to `'high'`** with a `debugLogger.warn` line (once per generator); if you want max effort, switch the baseURL to a DeepSeek-compatible endpoint that supports it.                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                             |
+| **Anthropic** (`api.deepseek.com/anthropic`)  | Same `output_config: { effort }` + beta header                                       | `'max'` is passed through unchanged.                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                        |
+| **Gemini** (`@google/genai`)                  | `thinkingConfig: { includeThoughts: true, thinkingLevel }`                           | `'low'` → `LOW`, `'high'`/`'max'` → `HIGH`, others → `THINKING_LEVEL_UNSPECIFIED` (Gemini has no `MAX` tier).                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                               |
 ### `reasoning: false`
-Setting `reasoning: false` (the literal boolean) explicitly disables thinking on every provider — useful for cheap side queries that don't benefit from reasoning. This is honored at the request level too via `request.config.thinkingConfig.includeThoughts: false` for one-off calls (e.g. suggestion generation).
+Setting `reasoning: false` (the literal boolean) explicitly disables thinking on models that support disabling — useful for cheap side queries that don't benefit from reasoning. This is honored at the request level too via `request.config.thinkingConfig.includeThoughts: false` for one-off calls (e.g. suggestion generation).
 On a `api.deepseek.com` baseURL, the OpenAI pipeline emits the explicit `thinking: { type: 'disabled' }` field that DeepSeek V4+ requires — the server-side default is `'enabled'`, so simply omitting `reasoning_effort` would still pay thinking latency/cost. Self-hosted DeepSeek backends (sglang/vllm) and other OpenAI-compatible servers do **not** receive this field; if you need to disable thinking on those, inject `thinking: { type: 'disabled' }` (or whatever knob your inference framework exposes) via `samplingParams`/`extra_body`.
-On an `openrouter.ai` baseURL, the OpenAI pipeline emits OpenRouter's provider-level `reasoning: { enabled: false }` field when reasoning is disabled. Other OpenAI-compatible servers do not receive this OpenRouter-specific field; use `samplingParams`/`extra_body` for their native disable knob.
+For known GPT models that allow disabling, non-OpenRouter endpoints receive `reasoning_effort: 'none'`; OpenRouter receives nested `reasoning: { enabled: false }` instead. Mandatory-thinking models reject off in model controls and omit unsupported disable values from requests, so `reasoning: false` cannot turn their thinking off. An explicit model reasoning capability takes precedence over the built-in tier list and selects the native disable field; OpenRouter keeps its provider-level disable behavior. The built-in mandatory set is `gpt-5`, `gpt-5-mini`, `gpt-5-nano`, `gpt-5-pro`, `gpt-5.1-codex`, `gpt-5.1-codex-max`, `gpt-5.2-codex`, `gpt-5.3-codex`, `gpt-5.2-pro`, `gpt-5.4-pro`, `gpt-5.5-pro`, and `gpt-6-astra`.
+On an `openrouter.ai` baseURL, the OpenAI pipeline emits OpenRouter's provider-level `reasoning: { enabled: false }` field when reasoning is disabled. Mandatory-thinking models do not receive this disable field. Other OpenAI-compatible servers do not automatically receive this OpenRouter-specific field; use their native disable knob.
 ### Interaction with `samplingParams` (OpenAI-compatible only)
 > [!warning]
 >
-> When `generationConfig.samplingParams` is set on an OpenAI-compatible provider, the pipeline ships those keys to the wire **verbatim** and skips the separate `reasoning` injection entirely. So a config like `{ samplingParams: { temperature: 0.5 }, reasoning: { effort: 'max' } }` will silently drop the reasoning field on OpenAI/DeepSeek requests. A `reasoning` object placed inside `samplingParams` is your own value and ships unchanged: the effort ceiling above applies only to the tier the pipeline injects from `/effort`.
+> Except for known GPT models and models with explicit reasoning capabilities, when `generationConfig.samplingParams` is set on an OpenAI-compatible provider, the pipeline ships those keys to the wire **verbatim** and skips the separate `reasoning` injection entirely. So a config like `{ samplingParams: { temperature: 0.5 }, reasoning: { effort: 'max' } }` will silently drop the reasoning field on OpenAI/DeepSeek requests. A `reasoning` object placed inside `samplingParams` is your own value and ships unchanged while reasoning is enabled: the effort ceiling above applies only to the tier the pipeline injects from `/effort`.
+>
+> Known GPT-5 models and GPT-6 Astra keep configured effort alongside unrelated sampling keys. For example, `{ samplingParams: { temperature: 0.5 }, reasoning: { effort: 'max' } }` sends `temperature: 0.5` and flat `reasoning_effort: 'xhigh'` on GPT-5.4, or `'max'` on GPT-5.6 / GPT-6 Astra. On an `openrouter.ai` baseURL the same clamped tier ships as nested `reasoning: { effort }` instead. On non-OpenRouter endpoints, explicit flat reasoning overrides win; nullish or empty-string flat placeholders allow the configured tier. On OpenRouter, a sampling flat override suppresses configured nested effort unless explicit model capabilities inject it; an extra-body-only flat override does not replace the configured nested effort.
 >
-> DashScope Qwen models are an exception: their provider reads `reasoning` directly and maps it to `reasoning_effort` or `enable_thinking`. On the qwen3.8-max family, provider-specific `samplingParams` fields still take precedence when the wire parameters conflict; on older qwen hybrids, a configured effort tier collapses to `enable_thinking: true`, which overrides a `samplingParams.enable_thinking` value.
+> DashScope Qwen models are another exception: their provider reads `reasoning` directly and maps it to `reasoning_effort` or `enable_thinking`. On the qwen3.8-max family, provider-specific `samplingParams` fields still take precedence when the wire parameters conflict; on older qwen hybrids, a configured effort tier collapses to `enable_thinking: true`, which overrides a `samplingParams.enable_thinking` value.
 >
-> If you set `samplingParams`, include the reasoning knob inside it directly — for DeepSeek that's `samplingParams.reasoning_effort`, for GPT-5/o-series it's `samplingParams.reasoning_effort` (their flat field) or `samplingParams.reasoning` (the nested object). For OpenRouter and other providers the field name varies; consult the provider docs.
+> For other models, include the provider's reasoning knob directly when using `samplingParams` — for DeepSeek that is `samplingParams.reasoning_effort`. Known GPT models map configured effort automatically; only add a raw override when intentionally bypassing that mapping. Raw nested `reasoning`, including `null`, remains a whole-object override while reasoning is enabled. Disabling via `reasoning: false` or request-level `includeThoughts: false` removes the nested value, including raw overrides in either layer. Non-OpenRouter GPT requests then send `reasoning_effort: 'none'` when disabling is allowed; OpenRouter uses its nested disable field instead. Mandatory-thinking models receive neither substitute. Outside OpenRouter its meaning depends on the gateway, so model controls show the model default. Any raw override that blocks a configured tier causes an explicit tier change to fail without saving a preference. The thinking switch restores configured raw defaults after disabling only when they permit thinking. If the raw state or configured reasoning default is off, the thinking switch cannot be turned on and the saved preference is retained. This also applies when explicit capabilities omit a default tier or expose only a thinking toggle. An explicit default command still resets the preference. Remove the blocking raw override to choose a different tier.
 >
 > The Anthropic and Gemini converters are unaffected — they always read `reasoning.effort` directly regardless of `samplingParams`.

package/bundled/qc-helper/docs/features/channels/dingtalk.md CHANGED Viewed

@@ -198,7 +198,7 @@ You can send photos and documents to the bot, not just text.
 **Files:** Send a PDF, code file, or any document. The bot downloads it from DingTalk's servers and saves it locally so the agent can read it with its file tools. Audio and video files are also supported. This works with any model.
-**Generated files:** Ask the agent explicitly to send a completed local file and it can return the file as a native DingTalk attachment. Files must be non-empty, no larger than 20 MB, and located inside the configured workspace or the system temporary directory. One response can send at most five files. Outbound file attachments are unavailable when `blockStreaming` is set to `"on"`; upload or delivery failures are reported in the final text instead.
+**Generated files:** Ask the agent explicitly to send a completed local file and it can return the file as a native DingTalk attachment. Files must be non-empty, no larger than 20 MB, and located inside the configured workspace or the system temporary directory. One response can send at most five files. Upload or delivery failures are reported in the final text instead.
 ## Forwarded Chat Records

package/bundled/qc-helper/docs/features/channels/github.md CHANGED Viewed

@@ -46,7 +46,6 @@ Add the channel to `~/.qwen/settings.json`:
       "allowedUsers": ["operator-github-username"],
       "sessionScope": "chat_thread",
       "cwd": "/path/to/your/project",
-      "blockStreaming": "off",
       "groupPolicy": "open",
       "groups": {
         "*": { "requireMention": true }
@@ -87,7 +86,6 @@ Local `gh` authentication requires an HTTPS `baseUrl` so the daemon host credent
 | `groupPolicy`             | `"disabled"`             | Must be `"open"`, `"allowlist"` with the repo (`owner/repo`) listed in `groups`, or `"pairing"` with the repo approved for notifications to flow |
 | `senderPolicy`            | `"allowlist"`            | Who can trigger the bot                                                                                                                          |
 | `groups.*.requireMention` | `true`                   | Require @mentions for ordinary comments; directed notification reasons still run                                                                 |
-| `blockStreaming`          | `"off"`                  | Always forced to `"off"`; intermediate model chunks aren't published; `"on"` is not supported                                                    |
 | `reasonFilter`            | unset                    | Optional allowlist of GitHub notification reasons to process                                                                                     |
 Use `reasonFilter` to drop noisy notification classes such as `ci_activity` or `state_change`. Do not use `reasonFilter: ["mention"]` as a replacement for `groups.*.requireMention`: GitHub's `mention` reason is sticky at the thread level, so real new @mentions can arrive later under `comment`, `subscribed`, `author`, or other reasons and would be skipped.
@@ -133,13 +131,7 @@ For an accepted issue or pull-request comment, the channel adds GitHub's `👀`
 ### Final-only output
-The GitHub channel always forces final-only delivery. The adapter sets `blockStreaming` to `"off"`, so intermediate model chunks are never published as separate comments and `blockStreaming: "on"` is not supported.
-```json
-{
-  "blockStreaming": "off"
-}
-```
+The GitHub channel publishes only completed responses. Intermediate model chunks are never published as separate comments.
 If GitHub returns a definite no-write delivery failure, such as a rate-limit
 response, the channel stores the final reply in

package/bundled/qc-helper/docs/features/channels/overview.md CHANGED Viewed

@@ -50,32 +50,29 @@ Channels are configured under the `channels` key in `settings.json`. Each channe
 ### Options
-| Option                   | Required         | Description                                                                                                                                                                                                             |
-| ------------------------ | ---------------- | ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
-| `type`                   | Yes              | Channel type: `telegram`, `weixin`, `qq`, `dingtalk`, `dws`, `wecom`, `feishu`, `github`, `gitlab`, or a custom type from an extension (see [Plugins](./plugins))                                                       |
-| `token`                  | Telegram         | Bot token. Supports `$ENV_VAR` syntax to read from environment variables. Not needed for WeChat, DingTalk, WeCom, or Feishu                                                                                             |
-| `clientId`               | DingTalk, Feishu | DingTalk AppKey or Feishu App ID. Supports `$ENV_VAR` syntax                                                                                                                                                            |
-| `clientSecret`           | DingTalk, Feishu | DingTalk AppSecret or Feishu App Secret. Supports `$ENV_VAR` syntax                                                                                                                                                     |
-| `botId`                  | WeCom            | WeCom intelligent robot Bot ID. Supports `$ENV_VAR` syntax. See [WeCom](./wecom)                                                                                                                                        |
-| `secret`                 | WeCom            | WeCom intelligent robot Secret. Supports `$ENV_VAR` syntax. See [WeCom](./wecom)                                                                                                                                        |
-| `model`                  | No               | Model to use for this channel (e.g., `qwen3.5-plus`). Overrides the default model. Useful for multimodal models that support image input                                                                                |
-| `senderPolicy`           | No               | Who can talk to the bot: `allowlist` (default), `open`, or `pairing`                                                                                                                                                    |
-| `allowedUsers`           | No               | List of user IDs allowed to use the bot (used by `allowlist` and `pairing` policies)                                                                                                                                    |
-| `sessionScope`           | No               | How sessions are scoped: `user` (default), `chat_thread`, or `single`. Legacy `thread` remains compatible when already configured but is not offered for new Web Shell configurations                                   |
-| `multiSession`           | No               | Retain up to eight owner-scoped named tasks in one chat. Requires daemon-managed mode, `sessionScope: "user"`, no webhooks or group-history backfill, and no enabled Channel loops                                      |
-| `messagePrefix`          | No               | Only dispatch user messages that begin with this exact, case-sensitive prefix after any leading `@mentions`; the prefix and following whitespace are removed before dispatch                                            |
-| `cwd`                    | No               | Working directory for the agent. Defaults to the current directory                                                                                                                                                      |
-| `approvalMode`           | No               | Tool approval mode for channel sessions. Unattended webhook tasks require `yolo`; the setting applies to every session on the channel                                                                                   |
-| `instructions`           | No               | Custom instructions prepended to the first message of each session                                                                                                                                                      |
-| `webhooks`               | No               | Webhook sources and delivery targets for daemon-managed channels. See [Webhook-triggered tasks](#webhook-triggered-tasks)                                                                                               |
-| `groupPolicy`            | No               | Group chat access: `disabled` (default), `allowlist`, `pairing`, or `open`. See [Group Chats](#group-chats)                                                                                                             |
-| `dmPolicy`               | No               | Private/DM access: `open` (default) or `disabled` (silently drop all DMs). Useful for group-only bots                                                                                                                   |
-| `groupHistoryLimit`      | No               | Opt-in group history backfill. `0` or omitted disables it. A positive number persists that many unmentioned group messages from authorized senders or members of approved paired groups for the next bot mention/reply. |
-| `groups`                 | No               | Per-group settings. Keys are group chat IDs or `"*"` for defaults. See [Group Chats](#group-chats)                                                                                                                      |
-| `dispatchMode`           | No               | What happens when you send a message while the bot is busy: `steer` (default), `collect`, or `followup`. See [Dispatch Modes](#dispatch-modes)                                                                          |
-| `blockStreaming`         | No               | Progressive response delivery: `on` or `off` (default). See [Block Streaming](#block-streaming)                                                                                                                         |
-| `blockStreamingChunk`    | No               | Chunk size bounds: `{ "minChars": 400, "maxChars": 1000 }`. See [Block Streaming](#block-streaming)                                                                                                                     |
-| `blockStreamingCoalesce` | No               | Idle flush: `{ "idleMs": 1500 }`. See [Block Streaming](#block-streaming)                                                                                                                                               |
+| Option              | Required         | Description                                                                                                                                                                                                             |
+| ------------------- | ---------------- | ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
+| `type`              | Yes              | Channel type: `telegram`, `weixin`, `qq`, `dingtalk`, `dws`, `wecom`, `feishu`, `github`, `gitlab`, or a custom type from an extension (see [Plugins](./plugins))                                                       |
+| `token`             | Telegram         | Bot token. Supports `$ENV_VAR` syntax to read from environment variables. Not needed for WeChat, DingTalk, WeCom, or Feishu                                                                                             |
+| `clientId`          | DingTalk, Feishu | DingTalk AppKey or Feishu App ID. Supports `$ENV_VAR` syntax                                                                                                                                                            |
+| `clientSecret`      | DingTalk, Feishu | DingTalk AppSecret or Feishu App Secret. Supports `$ENV_VAR` syntax                                                                                                                                                     |
+| `botId`             | WeCom            | WeCom intelligent robot Bot ID. Supports `$ENV_VAR` syntax. See [WeCom](./wecom)                                                                                                                                        |
+| `secret`            | WeCom            | WeCom intelligent robot Secret. Supports `$ENV_VAR` syntax. See [WeCom](./wecom)                                                                                                                                        |
+| `model`             | No               | Model to use for this channel (e.g., `qwen3.5-plus`). Overrides the default model. Useful for multimodal models that support image input                                                                                |
+| `senderPolicy`      | No               | Who can talk to the bot: `allowlist` (default), `open`, or `pairing`                                                                                                                                                    |
+| `allowedUsers`      | No               | List of user IDs allowed to use the bot (used by `allowlist` and `pairing` policies)                                                                                                                                    |
+| `sessionScope`      | No               | How sessions are scoped: `user` (default), `chat_thread`, or `single`. Legacy `thread` remains compatible when already configured but is not offered for new Web Shell configurations                                   |
+| `multiSession`      | No               | Retain up to eight owner-scoped named tasks in one chat. Requires daemon-managed mode, `sessionScope: "user"`, no webhooks or group-history backfill, and no enabled Channel loops                                      |
+| `messagePrefix`     | No               | Only dispatch user messages that begin with this exact, case-sensitive prefix after any leading `@mentions`; the prefix and following whitespace are removed before dispatch                                            |
+| `cwd`               | No               | Working directory for the agent. Defaults to the current directory                                                                                                                                                      |
+| `approvalMode`      | No               | Tool approval mode for channel sessions. Unattended webhook tasks require `yolo`; the setting applies to every session on the channel                                                                                   |
+| `instructions`      | No               | Custom instructions prepended to the first message of each session                                                                                                                                                      |
+| `webhooks`          | No               | Webhook sources and delivery targets for daemon-managed channels. See [Webhook-triggered tasks](#webhook-triggered-tasks)                                                                                               |
+| `groupPolicy`       | No               | Group chat access: `disabled` (default), `allowlist`, `pairing`, or `open`. See [Group Chats](#group-chats)                                                                                                             |
+| `dmPolicy`          | No               | Private/DM access: `open` (default) or `disabled` (silently drop all DMs). Useful for group-only bots                                                                                                                   |
+| `groupHistoryLimit` | No               | Opt-in group history backfill. `0` or omitted disables it. A positive number persists that many unmentioned group messages from authorized senders or members of approved paired groups for the next bot mention/reply. |
+| `groups`            | No               | Per-group settings. Keys are group chat IDs or `"*"` for defaults. See [Group Chats](#group-chats)                                                                                                                      |
+| `dispatchMode`      | No               | What happens when you send a message while the bot is busy: `steer` (default), `collect`, or `followup`. See [Dispatch Modes](#dispatch-modes)                                                                          |
 When `messagePrefix` is set, every user-authored message must begin with the prefix and a non-empty payload, for example `/review inspect #123`. Only the prefix and the mentions ahead of it are removed; a mention the user typed after the prefix reaches the agent unchanged. Shared and agent commands use the same rule (`/review /help`, `/review /clear`, and so on). Telegram's registered command-menu actions remain available without the prefix — unless the configured prefix is itself one of them, in which case the prefix wins and that command has to be sent prefixed too (`/new /new`). Attachments need a matching caption when the platform supports one; captionless Telegram, Feishu, WeChat, DingTalk and WeCom media messages continue to run, and their placeholder text is never quoted back as group history. Native todos, webhooks, and provider-generated assignment or review-request events also continue to run without a prefix because they are system events rather than chat messages.
@@ -417,33 +414,11 @@ You can also set dispatch mode per group, overriding the channel default:
 }
 ```
-## Block Streaming
+## Response delivery
-By default, the agent works for a while and then sends one large response. With block streaming enabled, the response arrives as multiple shorter messages while the agent is still working — similar to how ChatGPT or Claude show progressive output.
+Channels use their normal response delivery path. The shared delivery layer sends completed responses, and adapters may provide native progressive display, such as updating an interactive card in place. Platform message-length limits may still split long responses.
-```json
-{
-  "channels": {
-    "my-channel": {
-      "type": "telegram",
-      "blockStreaming": "on",
-      "blockStreamingChunk": { "minChars": 400, "maxChars": 1000 },
-      "blockStreamingCoalesce": { "idleMs": 1500 },
-      ...
-    }
-  }
-}
-```
-### How it works
-- The agent's response is split into blocks at paragraph boundaries and sent as separate messages
-- `minChars` (default 400) — don't send a block until it's at least this long, to avoid spamming tiny messages
-- `maxChars` (default 1000) — if a block gets this long without a natural break, send it anyway
-- `idleMs` (default 1500) — if the agent pauses (e.g., running a tool), send what's buffered so far
-- When the agent finishes, any remaining text is sent immediately
-Only `blockStreaming` is required. The chunk and coalesce settings are optional and have sensible defaults.
+The obsolete `blockStreaming`, `blockStreamingChunk`, and `blockStreamingCoalesce` settings are no longer supported and can be removed from channel configuration. They do not affect delivery. Channel settings management rejects newly added or changed values for these fields. An unchanged stored value is retained, or removed, when the edit keeps the channel's `type`; changing a channel's `type` requires removing these fields first.
 ## Scheduled Channel Loops

package/bundled/qc-helper/docs/features/channels/qqbot.md CHANGED Viewed

@@ -63,7 +63,6 @@ export QQ_APP_SECRET=<your-app-secret>
       "sessionScope": "user",
       "cwd": "/path/to/your/project",
       "instructions": "你是一个通过 QQ Bot 对话的 AI 助手。回复控制在 2000 字符以内。",
-      "blockStreaming": "on",
       "groupPolicy": "disabled",
       "groups": {
         "*": { "requireMention": true }
@@ -82,7 +81,7 @@ export QQ_APP_SECRET=<your-app-secret>
 | `sandbox`   | `false` | Set to `true` to use the QQ sandbox API environment (`sandbox.api.sgroup.qq.com`) |
 All standard channel options (see [Channel Overview](./overview#options)) are also supported:
-`senderPolicy`, `allowedUsers`, `sessionScope`, `cwd`, `instructions`, `groupPolicy`, `groups`, `dispatchMode`, `blockStreaming`, `blockStreamingChunk`, `blockStreamingCoalesce`.
+`senderPolicy`, `allowedUsers`, `sessionScope`, `cwd`, `instructions`, `groupPolicy`, `groups`, `dispatchMode`.
 ## Running

package/bundled/qc-helper/docs/features/commands.md CHANGED Viewed

@@ -871,14 +871,50 @@ session registry rather than deriving it.
 The `send_message` call only confirms the message was handed to the other
 session. What became of it arrives later as a receipt: if it was held,
-declined, refused, expired, or misaddressed (the address changed hands —
-list the agents again) — or released after a hold — a notice appears in
-the sending session's transcript (`Message to <name>: …`). Declined and
-refused are different answers: declined means someone reviewed the
-message and said no, while refused means that session's
-`agents.crossSessionInbound` is `refuse` and nobody saw it at all. The
-model that sent it is not told; if the other session replies, the reply
-arrives as a cross-session message.
+declined, refused, dropped, expired, or misaddressed (the address changed
+hands — list the agents again) — or released after a hold — a notice
+appears in the sending session's transcript (`Message to <name>: …`).
+Declined, refused and dropped are three different answers: declined means
+someone reviewed the message and said no, refused means that session's
+`agents.crossSessionInbound` is `refuse` and nobody saw it at all, and
+dropped means its inbox turned the message away before any of that (see
+below). The first drop is answered at once and the rest are folded into a
+receipt every few seconds, each naming the messages it stands for, so a
+run of them costs a handful of lines rather than one line each. The model that
+sent it is not told; if the other session replies, the reply arrives as a
+cross-session message.
+### Flood protection
+A session accepts up to 30 messages at once from one sender and then one
+every two seconds, and up to 32 at once from all senders together and
+then one a second. The second limit exists because a sender names itself:
+rotating that name gets a fresh allowance from the first limit but not
+from the second. It is barely above the first because every accepted
+message draws a receipt, and a session can only have so many of those
+going out at once. A message from another session that repeats that
+sender's previous message word for word within 30 seconds is also turned
+away — a model looping on one sentence mints a fresh message id every
+time, so the text is what catches it. Messages from a script the session
+started and from a trusted controller are exempt from the repeat check,
+because a hook reporting the same line twice is reporting two facts and a
+person saying "continue" twice means it twice; both are still subject to
+the rate limits. Finally, a message that is accepted but cannot be queued
+because the session already has 50 waiting is turned away too.
+A message turned away this way is never held, never shown to the model,
+and leaves no record, so the sender can try again later and land. The
+receiving session says so in its transcript at most once a minute per
+sender, with a count of what that line stands for. The sending session
+gets one receipt naming every message the burst cost it, and its
+transcript says to fold what still matters into one later message rather
+than re-sending.
+The sending side does not wait to find out. Each session tracks what it
+has sent to each address and refuses a send that the receiver would
+drop, so the model is told to batch before the message is written rather
+than after — and the receiver never spends a connection on a message it
+was going to turn away.
 ### Inbox authentication and scripted injection
@@ -904,7 +940,10 @@ recognized as the session's own rather than as another session's.
 Give every injection a fresh `msgId`. The receiving gate remembers the
 ids it has already settled, so a hook that reuses one is delivered the
-first time and silently deduplicated on every run after that.
+first time and silently deduplicated on every run after that. Repeating
+the same _text_ is fine — the repeat check above does not apply to a
+session's own processes — but the rate limits do apply, so a hook in a
+loop is dropped like any other flood.
 An injected message still goes through the inbound gate and is marked as
 not coming from the user, but the gate knows it came from the session's

package/bundled/qc-helper/docs/features/goals.md CHANGED Viewed

@@ -18,6 +18,8 @@ Creating, editing, or resuming a Goal requires a trusted workspace (`/trust`). H
 Once a Goal has billed a turn, the footer pill and every status card show what it has spent against the window it is allowed, as `1.2k/30.0m`. The figure counts the model calls the Goal makes in its own turns; subagents and the verifier's own checks are not included. The window is set by [`model.goalTokenBudget`](../configuration/settings.md); resuming a Goal that has spent its window grants another one on top of what it has already spent, so the figure reads `30.0m/60.0m` rather than starting over. A Goal with no budget shows only what it has spent. A Goal that has not billed a turn yet shows no figures at all.
+Each turn the session takes on its own reports what the Goal has spent so far, how many turns are behind it, and — unless the Goal runs unbounded — the window it is allowed. Every such turn except the final wind-down hand-off also carries standing instructions to re-check the workspace rather than trust earlier turns' reports, to work toward the end state the objective asks for, to do something different when the previous turn changed nothing (from the second turn on, once there is a previous turn to judge), and to check every requirement against citable evidence before proposing that the Goal is done.
 A long Goal periodically compresses the evidence it has recorded into checkpoint claims with a side model call, so later turns and the verifier still have it to cite. That call is bounded by [`model.goalCheckpointTimeoutSeconds`](../configuration/settings.md), 180 seconds by default; a checkpoint that does not finish in time is abandoned as an inconclusive check — the checkpoint stall streak is preserved rather than incremented — and a later turn retries it. The call is streamed, so the per-request transport timeout bounds only connect and first response, and the ceiling itself stops at the stream guards' 15-minute lifetime cap because past that the guard, not the setting, ends the call. That 15-minute limit on the setting is fixed, and raising the stream guard's own cap does not lift it.
 ## Interrupting a Goal

package/bundled/qc-helper/docs/features/sub-agents.md CHANGED Viewed

@@ -131,6 +131,14 @@ When a session is restored, compatible background agents are added back to the s
 Use continuation for related follow-up work. Launch a new agent when the task is unrelated or the previous agent cannot be resumed.
+## Notification Queue
+In the interactive TUI and ACP session, completion notifications from background agents, shells, monitors and workflows share a queue that drains into a model turn once the session is idle. These queues hold at most 20 notifications so a noisy producer cannot accumulate an unbounded backlog. The headless CLI's local queue is not capped by this rule.
+When a 21st notification arrives, Qwen Code evicts an interim monitor pulse first — the monitor's next poll supersedes it — and otherwise the oldest queued notification. Agent results, workflow results and scheduled prompts are never evicted in the interactive TUI; a notification that would displace one is dropped instead, and so is an arriving pulse when only terminal results are queued.
+Discarded notifications are reported rather than dropped quietly. The summary appears before the next notification in the live transcript. ACP also prefixes it to that turn's model input; the TUI keeps it parked for the next Notification batch so cron prompts still pass unchanged through slash, shell and `@` preprocessing. A daemon notification is recorded before it is acknowledged, so after a reload its durable record can precede the later overflow summary. ACP can discard a pending summary if the session is cleared or switched, or if a client cancels or preempts the notification turn. Discarding a notification never stops or deletes its task, and completed tasks retain their results; the summary points at `/tasks` and task output files when there is a task to inspect. A discarded scheduled prompt was never delivered and is not retried. A daemon notification that was recorded but could not be delivered live remains available in the session transcript and is reported separately from lost notifications.
 ## Agent Working Directory
 For a named regular subagent, `working_dir` pins the agent to an existing git worktree of the current repository. Relative paths resolve from the current directory, and the worktree must already be registered with git as a linked worktree of this repository.