@michengai/dsh-automation 0.1.54 → 0.1.58

package/CHANGELOG.md

@@ -6,6 +6,18 @@ Upcoming changes and recent published versions are listed below. Git tags and Gi
 
 ## Unreleased
 
+## 0.1.58 - 2026-10-07
+
+- Official Desktop can update this plugin in place. It no longer targets the web profile by mistake.
+- A new version is marked with a warning color. Version numbers keep their normal color.
+- After updating, fully quit and reopen DSH Desktop.
+
+## 0.1.55 - 2026-10-07
+
+- Agent automation create/update tools support provider, model, and reasoning effort, including resetting to the runtime global model or model default (#15).
+- Reject incomplete provider/model pairs and preserve explicit null on creation. Switching models clears an omitted reasoning override. `automation_list` now includes model IDs, supported reasoning efforts, the global model, and catalog failures.
+- Trim model IDs on save. Omitted model fields inherit a complete pair and never splice a session provider with the global model.
+
 ## 0.1.54 - 2026-10-06
 
 - The installed-plugin list shows “Automation” and a localized description instead of the package name. Titles and descriptions come from `meta.title` and `meta.description` in `locale/zh.json` and `locale/en.json`.

package/CHANGELOG.zh-CN.md

@@ -6,6 +6,18 @@
 
 ## 未发布
 
+## 0.1.58 - 2026-10-07
+
+- 官方桌面版可以在线更新这个插件，不再误装到 web profile。
+- 发现新版本时，提示文字会标成提醒色；版本号颜色不变。
+- 更新完成后需要完全退出并重新打开 DSH Desktop。
+
+## 0.1.55 - 2026-10-07
+
+- Agent 创建和更新定时任务支持提供方、模型和推理等级，也可恢复运行时全局模型或模型默认等级（#15）。
+- 拒绝不完整的 provider/model 配对，创建时保留显式 null。切换模型时清除未重新指定的旧推理等级。`automation_list` 返回模型 ID、支持的推理等级、全局模型和目录加载失败信息。
+- 保存模型 ID 时去掉首尾空白。省略模型时按完整配对继承，不把会话 provider 和全局 model 拼在一起。
+
 ## 0.1.54 - 2026-10-06
 
 - 已安装插件列表显示「Automation」和中文简介，不再只用包名。显示名来自 `locale/zh.json` 和 `locale/en.json` 的 `meta.title`、`meta.description`。

package/README.md

@@ -123,6 +123,10 @@ Open **Settings → Scheduled Tasks**, then use the panel as follows:
 
 Each dispatched run uses the saved prompt, workspace, model, and permission boundary. It does not reuse approvals from the source chat.
 
+`automation_create` and `automation_update` support the same model controls as Settings: `provider`, `model`, and `reasoning_effort`. Use provider/model IDs and a reasoning level published by that model in `reasoningEfforts`. On creation, omitted provider/model fields inherit the current session's complete pair; an incomplete session pair falls back to the whole global pair, never a mix of the two. Creation is rejected when neither source has a complete pair, instead of saving follow-global. Saved IDs are trimmed. Omitted reasoning effort uses the model default. On update, omitted fields keep their saved values. Set both `provider` and `model` to `null` to follow the global selection at each run, and set `reasoning_effort` to `null` to restore the model default. Create/update responses and `automation_list` include the saved target fields in the automation definition (`reasoningEffort` in the response).
+
+When specifying or updating a model, send both `provider` and `model` as non-empty IDs or both as `null`; incomplete pairs are rejected. Query `automation_list` first: `models` includes available IDs and `reasoning.efforts`, `defaultModel` is the global selection, and `modelFailures` reports catalog loading failures. Changing the saved provider/model pair clears the previous reasoning override unless you explicitly supply `reasoning_effort`; submitting the same pair preserves it. Model IDs and reasoning levels are not validated against the catalog when saving; catalog discovery lets the Agent choose from the Host's published options.
+
 ## Safety boundary
 
 | Item | Behavior |

package/README.zh-CN.md

@@ -123,6 +123,10 @@ dsh --profile web --dump-config
 
 每次派发都使用保存的任务说明、工作区、模型和权限边界，不会复用来源对话中的批准。
 
+`automation_create` 和 `automation_update` 支持与设置页相同的模型参数：`provider`、`model`、`reasoning_effort`。提供方和模型使用 ID，推理等级使用该模型公布的 `reasoningEfforts` 值。创建时省略 provider/model 会继承当前会话的完整配对；会话配对不完整时整对回退全局，不会把两边拼在一起。两边都没有完整配对时拒绝创建，而不是写成跟随全局。保存时去掉 ID 首尾空白。省略推理等级使用模型默认。更新时省略字段保持原值。将 `provider` 和 `model` 都设为 `null`，任务就会在每次运行时跟随当时的全局选择；`reasoning_effort=null` 清除固定推理等级，恢复模型默认。创建、更新和 `automation_list` 的返回定义包含保存的目标字段（返回中的推理等级字段为 `reasoningEffort`）。
+
+指定或修改模型时，必须成对传入非空的 `provider` 和 `model` ID，或两项都设为 `null`，不完整配对会被拒绝。先查询 `automation_list`：`models` 包含可选 ID 和 `reasoning.efforts`，`defaultModel` 是全局选择，`modelFailures` 提示目录加载失败。切换保存的模型配对时，若没有显式传入 `reasoning_effort`，会清除旧等级；重复提交相同配对则保留。保存时不会依据目录校验模型 ID 和推理等级；Agent 可通过目录查询选择宿主公布的选项。
+
 ## 权限与安全边界
 
 | 项目 | 行为 |