dsh-gateway-provider 1.1.2 → 1.1.3

package/README.md

@@ -66,6 +66,48 @@ it should answer `✓ Connected — N models`. Not using the public newapi
 cloud? Set **Base URL** on the card (or `baseURL` in config) to your own
 gateway address first.
 
+## Terminal / headless
+
+The host half of this plugin (the `llm-newapi` provider route) works in any
+profile — including terminal / TUI setups such as
+[dsh-tui](https://github.com/ccch1mneyyy/dsh-TUI). **Settings → Gateway
+Models is web-only**; without the web UI you configure the gateway by
+overriding the bundle's `llm-newapi` entry in the profile patch. A patch
+layer replaces `config` wholesale, so write the whole block out:
+
+```yaml
+# ~/.dsh/profiles/<name>/cordis.patch.yml
+# Override by id — do not insert a second llm-newapi row (duplicate = loader error).
+- id: llm-newapi
+  config:
+    label: My Gateway
+    baseURL: https://gw.example.com
+    apiKeyEnv: GW_API_KEY
+    models:
+      - id: some-model
+        protocol: gemini
+```
+
+Store the key under the name in `apiKeyEnv` (same rules as Install):
+
+```sh
+echo "GW_API_KEY: sk-REPLACE_WITH_YOUR_KEY" >> ~/.dsh/.credentials.yaml
+# or: export GW_API_KEY=sk-REPLACE_WITH_YOUR_KEY
+```
+
+Install into a non-web profile the same way — swap the profile name:
+
+```sh
+dsh plugin --profile dsh-tui add dsh-gateway-provider
+```
+
+After a restart the gateway models should appear in the TUI model picker.
+Connection Test / Sync / per-model editing still need the web settings page
+(or hand-edits of the YAML above). A future in-terminal settings surface
+(`tuiSettingsSections`, see [#5](https://github.com/Luck9Star/dsh-gateway-provider/issues/5))
+may cover the scalar fields; until then the YAML override is the supported
+path.
+
 ## Daily use
 
 Everything lives in **Settings → Gateway Models**:

package/docs/README.zh.md

@@ -42,6 +42,46 @@ dsh --profile web
 
 **预期结果：** 模型选择器多出一条 "NewAPI" 路由，列出网关的对话模型，最新在前。在网关卡片上点 **Test** —— 应显示 `✓ Connected — N models`。用的不是 newapi 公有云？先在卡片上改 **Base URL**（或配置里的 `baseURL`）指向你自己的网关地址。
 
+## 终端 / 无界面（headless）
+
+本插件的 host 半边（`llm-newapi` provider 路由）在任意 profile 都能用 ——
+包括终端 / TUI（例如 [dsh-tui](https://github.com/ccch1mneyyy/dsh-TUI)）。
+**Settings → Gateway Models 仅限 web**；没有 web UI 时，通过在 profile
+patch 里覆盖 bundle 插入的 `llm-newapi` 条目来配置网关。patch 层会整段
+替换 `config`，所以要把整个块写全：
+
+```yaml
+# ~/.dsh/profiles/<name>/cordis.patch.yml
+# 按 id 覆盖 —— 不要再 insert 第二条 llm-newapi（重复行 = 加载器报错）。
+- id: llm-newapi
+  config:
+    label: My Gateway
+    baseURL: https://gw.example.com
+    apiKeyEnv: GW_API_KEY
+    models:
+      - id: some-model
+        protocol: gemini
+```
+
+密钥按 `apiKeyEnv` 里的名字存放（规则与「安装」一节相同）：
+
+```sh
+echo "GW_API_KEY: sk-REPLACE_WITH_YOUR_KEY" >> ~/.dsh/.credentials.yaml
+# 或: export GW_API_KEY=sk-REPLACE_WITH_YOUR_KEY
+```
+
+装到非 web profile 时同样操作，换成对应 profile 名即可：
+
+```sh
+dsh plugin --profile dsh-tui add dsh-gateway-provider
+```
+
+重启后，网关模型会出现在 TUI 的模型选择器里。连接测试 / 同步 /
+按模型编辑仍需要 web 设置页（或手工改上面的 YAML）。之后可能会加终端内
+设置面（`tuiSettingsSections`，见
+[#5](https://github.com/Luck9Star/dsh-gateway-provider/issues/5)）覆盖标量
+字段；在那之前，YAML 覆盖就是受支持的用法。
+
 ## 日常使用
 
 一切都在 **Settings → Gateway Models**：

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "dsh-gateway-provider",
-  "version": "1.1.2",
+  "version": "1.1.3",
   "description": "Generic LLM gateway model provider plugin for DeepSeek Harness: mounts newapi / LiteLLM / Higress / any OpenAI-compatible gateway as provider routes, auto-discovers the model list, enriches every model with models.dev parameters, and dispatches each model over its native wire protocol (OpenAI / Anthropic / Gemini) via the pi-ai SDK",
   "repository": {
     "type": "git",
@@ -70,10 +70,10 @@
   },
   "peerDependencies": {
     "@deepseek-ai/cordis": "^4.0.2",
-    "@deepseek-ai/dsh-credentials": "0.2.1-rc.2",
-    "@deepseek-ai/dsh-launch-environment": "0.2.1-rc.2",
-    "@deepseek-ai/dsh-llm": "0.2.1-rc.2",
-    "@deepseek-ai/dsh-timeout": "0.2.1-rc.2",
+    "@deepseek-ai/dsh-credentials": ">=0.2.1-rc.2 <0.2.2-0",
+    "@deepseek-ai/dsh-launch-environment": ">=0.2.1-rc.2 <0.2.2-0",
+    "@deepseek-ai/dsh-llm": ">=0.2.1-rc.2 <0.2.2-0",
+    "@deepseek-ai/dsh-timeout": ">=0.2.1-rc.2 <0.2.2-0",
     "@deepseek-ai/schemastery": "^3.18.1"
   }
 }