DeepSeek Harness 플러그인 — 816 페이지

独立模型数据源与官方清单发现，基于证据规划 pi-ai 补缺、待确认与宽限移除；可恢复目录事务、内容与修订所有权校验，附 provider 设置及监控。

Model relay station for DeepSeek Harness (模型中转站): serve any registered DSH LLM provider (e.g. @tnnevol/dsh-codebuddy) over an OpenAI-compatible /v1 API, so other projects can call those models with a standard OpenAI client.

Max-retries control for DeepSeek Harness: adds one preference card to the web client's Settings → Model page that sets how many times a failed model request is retried.

DeepSeek Harness plugin: role-based model routing — the planner agent runs on deepseek-v4-pro, delegated executor subagents run on deepseek-v4-flash.

Unified ModelID routing for DeepSeek Harness: one logical model id over multiple providers, first-token failover with cooldown, health-aware ranking, fallback reasoning efforts, purpose-based simple/complex task split, per-tier reasoning effort, and a settings-page management panel.

DSH 插件：为模型选择器添加跨平台搜索过滤功能，支持商汤、百炼等所有配置的模型源。

Adds a search box to the DSH Web GUI model switcher (the composer model seat), filtering the provider-grouped model list by model name / id / provider while preserving the official selection behavior and reasoning-effort picker.

Searchable model picker for the DSH web composer — replaces the stock conversation.input.model seat with a filter-as-you-type popover.

DSH client plugin — keyword search when selecting models in the Web GUI

DSH web UI 插件：把输入框官方「模型选择」控件替换为两个独立按钮（供应商选择 + 模型选择，两级联动），保留推理等级调节；只动这一处控件，不改 DSH 源码。

DeepSeek Harness model selector upgrade: one-level model & reasoning-effort picker with 5 Liang tiers, custom model aliases, and peak/off-peak price countdown.

Model Shelf for DSH: a wide, collapsible model picker with a manually organized uncommon list and explicit provider/model identity.

Composer model selector shows Provider/Model — 对话框模型选择器显示「供应商/模型」

DSH Web UI model seat: 3-level quick slider (low/medium/high, each preconfigured with model + reasoning effort), advanced model/effort picker, modality icons, and per-provider usage/balance display. DSH 模型选择器增强：低中高三档快切滑块 + 高级选择 + 模态图标 + 供应商用量显示。

Smart model setup for DeepSeek Harness third-party API providers: recognize a model ID and fill in context window, input modalities and reasoning levels.

DeepSeek Harness plugin: shows model usage and account balance in the session header — OpenCode Go plan usage (rolling / weekly / monthly) with the selected model's monthly cap, and DeepSeek account balance — each only while a matching model is the active selection.

按键切换模型（DSH 全局插件）：把任意按键/组合键映射到模型与思考强度，一键切换。

统一管理 DSH 里的所有模型：设置页「模型开关」列出全部 provider 与模型（含插件动态注册的，内置 Models 页看不见的那批），支持逐模型与整组开关；关闭后模型会从所有模型选择器中消失。另提供 list_models 工具做只读查看。

Visual custom-provider presets and model switching for DeepSeek Harness Web.

One-click model + reasoning-effort switcher for the DeepSeek Harness web UI: capsule trigger, gradient effort slider with detents, and a premium MAX-state visual.

DeepSeek Harness plugin: per-provider and per-model switches for the composer's model picker. Settings → Models gains a toggle on every provider card and inside every provider's model catalog; switched-off entries stop being offered, without touching provider config or interrupting running sessions.

DSH 模型插件：监控各服务商上游 /models，列出新增模型并勾选添加（检查只读、显式应用、不自动删除），并为 OpenCode Go 路由按会话附带 x-opencode-session。Monitor provider model catalogs, add newly released models to DeepSeek Harness, and fix OpenCode Go's 400 MissingSessionID.

Sync live provider model lists into DSH settings and show 5h/7d quota rings for the current session model.

Auto-populate and auto-update the list of llm-pi-ai providers' models. Turn the plugin on and off to update on-demand