dsh-packycode-api 1.1.5 → 1.1.7

package/CHANGELOG.md

@@ -1,5 +1,18 @@
 # Changelog
 
+## 1.1.7
+
+- Fixed HTTP 405 from the settings RPC after a first hot install. The shipped Connection getter retains its own row's context, which need not inject `webServer`; the browser channel now registers with this plugin's `connection` / `webServer` injection scope as its explicit owner. Native Connection admission, authentication and RPC envelopes remain in use. The explicit-owner entry is internal to the shipped service, so other Connection implementations retain the public `rpc.handle` contract.
+- Removed the bundle override of the protected `connection` row. First installation now mounts the plugin's own channel without rebuilding that row or restarting DSH; unloading or losing the optional webserver scope removes the channel. Installation instructions now describe the restart-free first install.
+- Added real Connection / Cordis HTTP regressions for the 0.1.5 and 0.2.0-rc.2 hosts, covering first install, authenticated settings access, uninstall/reinstall and optional webserver availability. The 0.2 test host uses a separate development fixture with matching host dependencies.
+- Validation: 284 tests passed, both bundles rebuilt and the tarball generated. The six HTTP lifecycle/authentication checks also passed against the built host bundle, without restarting the fixture's existing server or Connection service.
+
+## 1.1.6
+
+- Fixed the missing `image` input capability on `deepseek-flash`, following [DeepSeek's official vision guide](https://api-docs.deepseek.com/guides/vision/). DSH can now send image attachments through the existing Chat Completions route. Audited the two GPT models, Grok 4.7 and three Claude models against their vendor documentation; all six already declare image input.
+- Added a regression that sends an image through the registered host route and checks the actual Chat Completions `image_url` payload. `scripts/live-check.mjs --with-vision` also checks DeepSeek's answers to two color fixtures without revealing the colors in the prompt.
+- Validation: 278 offline tests, both rebuilt bundles and the versioned tarball. The built route passed all four HTTP model checks, key rotation and secret-free RPC projections. DeepSeek correctly identified both 512×512 red and blue PNGs through the attachment service seam.
+
 ## 1.1.5
 
 - Added `grok-4.7`, displayed as **Grok 4.7**, to the PackyCode `openai-responses` route at `https://cf.api.fan/v1`. [The official Grok 4.7 guide](https://docs.x.ai/developers/grok-4-7) declares a 500,000-token context, text/image input, text output, and `low` / `medium` / `high` / `xhigh` reasoning (default `high`, no off or max). There is no separate documented text-output cap; the numeric pi-ai ceiling is 500,000, bounded by remaining context per request. Only GPT Responses requests keep `prompt_cache_key`: the gateway's Grok route rejects this field with a misleading `Incorrect API key provided` error. Non-GPT Responses requests use pi-ai's `cacheRetention: 'none'` to omit it; the model, protocol, effort and plugin-owned credential stay the same.

package/README.en.md

@@ -45,9 +45,9 @@ Open the desktop application's **Plugins** page, click **Add plugin**, and enter
 dsh-packycode-api
 ```
 
-Choose the **official** install source, then click **Install**. To pin a version, enter `dsh-packycode-api@1.1.5`.
+Choose the **official** install source, then click **Install**. To pin a version, enter `dsh-packycode-api@1.1.7`.
 
-Then **manually restart the desktop application**. Refreshing the page does not reload the plugin inside the Host process.
+A **first install requires no DSH restart**: open **Settings → PackyCode** and add a key immediately. When updating an already loaded older version, follow the restart guidance in the **Plugin version** card.
 
 ### 2. Add a key
 
@@ -64,7 +64,7 @@ Choose a model from the `PackyCode` group. `deepseek-flash`, `gpt-6.1-sol`, `gpt
 
 The **Plugins** page shows the installed state and version, and is also where you enable, disable, or uninstall.
 
-**For non-desktop profiles only:** `dsh plugin --profile <name> add dsh-packycode-api@1.1.5`. Do not use it for `desktop` — the CLI refuses it (`profile "desktop" is managed exclusively by the Electron application`).
+**For non-desktop profiles only:** `dsh plugin --profile <name> add dsh-packycode-api@1.1.7`. Do not use it for `desktop` — the CLI refuses it (`profile "desktop" is managed exclusively by the Electron application`).
 
 A local checkout can be linked with `dsh plugin --profile <name> add link:/path/to/dsh-packycode-api`.
 
@@ -138,7 +138,7 @@ Claude Code CLI → temporary local MCP web_search → selected PackyCode DeepSe
 
 | Model | How it runs | Effort | Token group |
 | --- | --- | --- | --- |
-| `deepseek-flash` | HTTP, chat/completions | Can be off; `low` / `high` / `max` | `deepseek-officially`, `deepseek-sale` |
+| `deepseek-flash` | HTTP, chat/completions, images allowed | Can be off; `low` / `high` / `max` | `deepseek-officially`, `deepseek-sale` |
 | `gpt-6.1-sol` | HTTP, Responses, images allowed | Cannot be off; `low` / `medium` / `high` / `xhigh` / `max` | `codex`, `codex-sale` |
 | `gpt-6-astra` | Same | Same | Same |
 | `grok-4.7` (Grok 4.7) | HTTP, Responses, text/image input; 500K context | Cannot be off; `low` / `medium` / `high` / `xhigh` | The Grok coverage read from the gateway in Settings |
@@ -148,6 +148,8 @@ Claude Code CLI → temporary local MCP web_search → selected PackyCode DeepSe
 
 A model in the list does not mean the current token group allows it. When changing groups, do not drop the group that serves the model you still want.
 
+All seven models in the current catalog declare text and image input. DeepSeek Flash follows the [official vision guide](https://api-docs.deepseek.com/guides/vision/); images are read through DSH's attachment service and sent with the Chat Completions request.
+
 Grok 4.7's context and reasoning levels follow the [official guide](https://docs.x.ai/developers/grok-4-7). There is no separate text-output limit; the plugin uses the context size as its numeric output ceiling, bounded by the remaining context per request.
 
 Only GPT Responses requests keep `prompt_cache_key`; other Responses models omit it by default. This gateway's Grok route reports a misleading `Incorrect API key provided` error when the field is present.

package/README.md

@@ -49,9 +49,9 @@
 dsh-packycode-api
 ```
 
-在**安装源**中选择**官方源**，然后点击**安装**。需要固定版本时填入 `dsh-packycode-api@1.1.5`。
+在**安装源**中选择**官方源**，然后点击**安装**。需要固定版本时填入 `dsh-packycode-api@1.1.7`。
 
-完成后**手动重启桌面应用**。仅刷新页面不会重新加载 Host 里的插件。
+首次安装后**无需重启 DSH**，可直接打开**设置 → PackyCode** 添加 Key。更新已加载的旧版时，按「插件版本」卡片的重启提示操作。
 
 ### 2. 添加 Key
 
@@ -68,7 +68,7 @@ dsh-packycode-api
 
 桌面应用的**插件**页可以看到已安装状态与版本，也可以在那里启用、停用或卸载。
 
-**仅用于非 desktop profile：** `dsh plugin --profile <name> add dsh-packycode-api@1.1.5`。不要对 `desktop` 使用它 —— CLI 会直接拒绝（`profile "desktop" is managed exclusively by the Electron application`）。
+**仅用于非 desktop profile：** `dsh plugin --profile <name> add dsh-packycode-api@1.1.7`。不要对 `desktop` 使用它 —— CLI 会直接拒绝（`profile "desktop" is managed exclusively by the Electron application`）。
 
 本地开发可以用绝对路径：`dsh plugin --profile <name> add link:/path/to/dsh-packycode-api`。
 
@@ -142,7 +142,7 @@ Claude Code CLI → 本机临时 MCP web_search → 插件选定的 DeepSeek/GPT
 
 | 模型 | 怎么调用 | 推理档位 | 适合的令牌分组 |
 | --- | --- | --- | --- |
-| `deepseek-flash` | HTTP，chat/completions | 可关；`low` / `high` / `max` | `deepseek-officially`、`deepseek-sale` |
+| `deepseek-flash` | HTTP，chat/completions，可带图 | 可关；`low` / `high` / `max` | `deepseek-officially`、`deepseek-sale` |
 | `gpt-6.1-sol` | HTTP，Responses，可带图 | 不能关；`low` / `medium` / `high` / `xhigh` / `max` | `codex`、`codex-sale` |
 | `gpt-6-astra` | 同上 | 同上 | 同上 |
 | `grok-4.7`（Grok 4.7） | HTTP，Responses，可带图；500k 上下文 | 不能关；`low` / `medium` / `high` / `xhigh` | 以设置页从网关读取的 Grok 分组覆盖为准 |
@@ -152,6 +152,8 @@ Claude Code CLI → 本机临时 MCP web_search → 插件选定的 DeepSeek/GPT
 
 列表里看得到模型，不代表当前令牌分组一定放行。换分组时不要把对应分组一起去掉。
 
+当前目录中的 7 个模型都声明了文字和图像输入。DeepSeek Flash 的识图能力按 [官方图像理解文档](https://api-docs.deepseek.com/zh-cn/guides/vision/) 配置；图片通过 DSH 附件服务读取，再随 chat/completions 请求发送。
+
 Grok 4.7 的上下文与推理档位来自 [官方文档](https://docs.x.ai/developers/grok-4-7)。官方没有单独限制文字输出长度；插件将输出上限设为上下文大小，并按每次请求的剩余上下文收紧。
 
 Responses 请求只有 GPT 模型保留 `prompt_cache_key`，其他 Responses 模型默认省略。当前网关的 Grok 路由会将这个字段误报成 `Incorrect API key provided`。

package/cordis.patch.yml

@@ -5,19 +5,3 @@
 - insert:
     - id: packycode-api
       name: 'dsh-packycode-api'
-
-# Connection registers a plugin's RPC channel through `owner.webServer`, where
-# the owner is the Connection row's own fiber — not the caller's. The shipped
-# dsh-web-app row injects `webRuntime` alone, so that read is refused with
-# `cannot get property "webServer" without inject`, the channel never reaches
-# the web server, and the browser's calls answer HTTP 405 while the boot log
-# stays clean.
-#
-# An entry patch replaces the whole inject list, so name every service that
-# fiber actually touches: `credentials` is the module's own inject, `webRuntime`
-# carries the row's `trustedHosts: !!js ctx.webRuntime.trustedHosts`, and
-# `webServer` is what channel registration reads. The name guard leaves a
-# replacement Connection implementation untouched.
-- id: connection
-  name: '@deepseek-ai/dsh-client-connection'
-  inject: [credentials, webRuntime, webServer]

package/lib/index.js

@@ -600,7 +600,7 @@ const PACKYCODE_MODELS = Object.freeze([
 		provider: PROVIDER,
 		baseUrl: COMPLETIONS_BASE,
 		reasoning: true,
-		input: Object.freeze(["text"]),
+		input: Object.freeze(["text", "image"]),
 		cost: NO_COST,
 		contextWindow: 1e6,
 		maxTokens: 384e3,
@@ -6915,15 +6915,16 @@ function apply(ctx, config) {
 		modelSelection
 	});
 	/**
-	* Connection registers a channel through `owner.webServer`, read on its own
-	* row's fiber — which is why this plugin ships the `connection` inject patch
-	* in `cordis.patch.yml`. Without that service the read is refused, the
-	* channel never reaches the web server, and the browser answers HTTP 405
-	* while the boot log stays clean. A profile with neither service (a headless
-	* run) simply never runs this callback.
+	* The shipped Connection getter retains its own service context, whose row
+	* need not inject webServer. Register with our injected context explicitly:
+	* a first hot install then mounts the channel without rebuilding Connection.
+	* The native registry still owns admission and the wire protocol. Its
+	* explicit-owner entry is internal, so retain the public RPC contract for
+	* other Connection implementations. A headless profile can omit this scope.
 	*/
 	ctx.inject(["connection", "webServer"], (connectionCtx) => {
-		ctx.effect(() => connectionCtx.connection.rpc.handle(CHANNEL, handler), "packycode-api: loopback account RPC");
+		const service = connectionCtx.connection;
+		connectionCtx.effect(() => typeof service.register === "function" ? service.register(connectionCtx, CHANNEL, handler) : service.rpc.handle(CHANNEL, handler), "packycode-api: loopback account RPC");
 	});
 	ctx.inject(["tools"], (toolsCtx) => {
 		ctx.effect(() => toolsCtx.tools.register(createClaudeCodeTool({

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "dsh-packycode-api",
-  "version": "1.1.5",
+  "version": "1.1.7",
   "description": "Use a PackyCode (PackyAPI) API key in DeepSeek Harness through a built-in multi-protocol model route with key rotation and balance display",
   "type": "module",
   "main": "./lib/index.js",
@@ -82,13 +82,18 @@
   },
   "devDependencies": {
     "@deepseek-ai/cordis": "4.0.2",
+    "@deepseek-ai/dsh-brand": "0.1.5-rc.3",
+    "@deepseek-ai/dsh-client-connection": "0.1.5-rc.3",
     "@deepseek-ai/dsh-credentials": "0.1.5-rc.2",
+    "@deepseek-ai/dsh-invariants": "0.1.5-rc.3",
     "@deepseek-ai/dsh-llm": "0.1.5-rc.2",
     "@deepseek-ai/dsh-llm-pi-ai": "0.1.5-rc.2",
+    "@deepseek-ai/dsh-scope": "0.1.5-rc.3",
     "@deepseek-ai/dsh-settings": "0.1.5-rc.2",
     "@deepseek-ai/dsh-tools": "0.1.5-rc.2",
     "@deepseek-ai/schemastery": "3.18.4",
     "@earendil-works/pi-ai": "0.85.1",
+    "packycode-host-v020-fixture": "file:tests/fixtures/host-v020",
     "react": "18.3.1",
     "tsdown": "0.22.2",
     "yaml": "2.9.1"