dsh-gateway-provider 1.0.5 → 1.0.6

package/README.md

@@ -131,12 +131,24 @@ the bridge is lifted from the official `dsh-llm-pi-ai` adapter.
 ```sh
 git clone https://github.com/Luck9Star/dsh-gateway-provider
 cd dsh-gateway-provider
-npm run link              # symlink into your dsh profile (single instanceof safety)
+npm install                # installs the plugin-pinned pi-ai copy (direct dependency)
+npm run link               # symlink @deepseek-ai/* into your dsh profile (single instanceof safety)
 npm run test:client       # settings-UI render, both locales
 npm run test:urls         # URL/derivation units
 npm run smoke             # live gateway round-trip (needs a real key)
 ```
 
+### Why the plugin pins its own pi-ai
+
+`@earendil-works/pi-ai` is a direct dependency (exact-pinned), independent of
+the pi-ai version bundled with the harness. This decouples the gateway model
+catalog (thinking levels, per-provider compat such as zhipu GLM's
+`supportsDeveloperRole: false`) from harness upgrades: a model missing from
+the harness's older catalog no longer degrades request encoding. The
+plugin↔harness boundary passes plain data (`GenerateOptions` in, dsh
+`StreamChunk`s out; `lib/pi-bridge.js` never leaks pi-ai objects across), so
+the plugin's pi-ai copy and the harness's own coexist safely in one process.
+
 Developing from a checkout: point the profile's `package.json` at
 `"dsh-gateway-provider": "link:/abs/path"` and re-run `pnpm install` in the
 profile. Do **not** also add an `id: llm-newapi` row to the profile's own

package/docs/README.zh.md

@@ -91,12 +91,17 @@ dsh --profile web
 ```sh
 git clone https://github.com/Luck9Star/dsh-gateway-provider
 cd dsh-gateway-provider
-npm run link              # 软链进 dsh profile（保证单实例）
+npm install               # 安装插件自带的 pi-ai 固定副本（直接依赖）
+npm run link              # @deepseek-ai/* 软链进 dsh profile（保证单实例）
 npm run test:client       # 设置页渲染测试，双语言
 npm run test:urls         # URL 派生单元测试
 npm run smoke             # 真实网关往返（需要真 key）
 ```
 
+### 为什么插件自带 pi-ai
+
+`@earendil-works/pi-ai` 是精确锁定的直接依赖，与 harness 自带的 pi-ai 版本解耦：网关模型目录（思考档位、各家 compat，如智谱 GLM 的 `supportsDeveloperRole: false`）不再随 harness 升级被动漂移——harness 旧目录里缺的模型不会再退化请求编码。插件与 harness 的边界只传纯数据（`GenerateOptions` 进、dsh `StreamChunk` 出，`lib/pi-bridge.js` 从不把 pi-ai 对象泄漏到边界外），因此插件副本与 harness 自带副本可在同一进程内安全共存。
+
 从本地检出开发：把 profile 的 `package.json` 指向
 `"dsh-gateway-provider": "link:/绝对/路径"`，然后在 profile 里重跑 `pnpm install`。**不要**再往 profile 自己的 `cordis.patch.yml` 里加 `id: llm-newapi` 行 —— bundle patch 已提供（重复行 = 加载器报错）。
 

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "dsh-gateway-provider",
-  "version": "1.0.5",
+  "version": "1.0.6",
   "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",
@@ -61,7 +61,9 @@
       "patch": "./cordis.patch.yml"
     }
   },
-  "dependencies": {},
+  "dependencies": {
+    "@earendil-works/pi-ai": "0.84.4"
+  },
   "peerDependencies": {
     "@deepseek-ai/cordis": "^4.0.1",
     "@deepseek-ai/dsh-credentials": "^0.1.0-rc.7",
@@ -69,7 +71,6 @@
     "@deepseek-ai/dsh-llm": "^0.1.0-rc.7",
     "@deepseek-ai/dsh-settings": "^0.1.0-rc.7",
     "@deepseek-ai/dsh-timeout": "^0.1.0-rc.7",
-    "@deepseek-ai/schemastery": "^3.18.1",
-    "@earendil-works/pi-ai": "^0.82.1"
+    "@deepseek-ai/schemastery": "^3.18.1"
   }
 }