890 results
Reasoning effort slider for DeepSeek Harness — a client (web) bundle that injects a per-session slider into the conversation input, shown when the selected model declares reasoning effort levels. Works with any provider that has reasoning efforts (Bailian, DeepSeek official, OpenAI o-series, etc.)
Subscription sign-in console for DeepSeek Harness: sign in to ChatGPT (Codex), Claude, GitHub Copilot and OpenRouter through DSH's own authorization flows, from one page that lists every registered flow instead of one provider.
DeepSeek Harness WSL plugin: call TypeSafe Jev / OpenRouter System One for noul, choice, and score decisions (jev_ask / jev_check / jev_rank).
Device-workspace federation for the dsh web GUI: pair into another DSH instance and browse its projects and sessions from the local sidebar, with each peer session opening in the local center column through a loopback embed proxy
A Git worktree plugin for DeepSeek Harness that creates task branches, registers worktrees as DSH Workspaces, and opens isolated sessions.
Eyes for text-only DeepSeek: view_image tool (any OpenAI-compatible VLM, local Ollama or cloud) + chat image-attachment bridge that rewrites pasted/dropped images into view_image path markers
Question index for DeepSeek Harness Web UI: an ordered, collapsible list of your asked questions floating on the right side of the conversation; click one to jump the session to that point. Auto-indexes the full history of every session on open and persists it locally.
DSH 对话内联生图插件：OpenAI 兼容接口，图片以 generated-image 块直接显示在对话框内，设置项在 设置→插件→可配置。
DSH model management, synchronized model visibility, and OpenAI Responses web search.
DSH web plugin: Codex-style /side side conversations. Open a temporary fork of the current chat in a floating panel — ask side questions without interrupting the main task or polluting its context.
Multi-channel decision & remote-control bridge for DeepSeek Harness: pushes any workspace's approval/question cards to IM channels (WeChat iLink first) and answers them from chat, with a session console, per-channel security, and an open push service.
Open the DSH Web GUI in an app-style WebView2 window on profile start, with a settings card for browser paths
Add an "Open folder" entry to the workspace session row overflow (⋯) menu; opens the session's workspace folder with the host OS.
ChatGPT OAuth and Codex models for DeepSeek Harness.
ChatGPT Plus/Pro OAuth login for DeepSeek Harness openai-codex models
Context Guardian as a native DeepSeek Harness (DSH) bundle: deterministic, seq-pointer-preserving compaction with model-facing recall and search. Companion to dsh-tool-guardian.
DSH 悬浮 API 余额徽章：可拖动、可缩放、亚克力质感、文字随背景明暗自适应。支持 DeepSeek / Moonshot / OpenAI / 自定义接口。· A floating API-balance badge for DeepSeek Harness: draggable, resizable, acrylic, adaptive text color.
Vendor logo seat in the DSH web composer: shows which provider serves the current model (DeepSeek / OpenAI / Anthropic / OpenCode / ...), live from the official per-session ModelDirectory.
DeepSeek Harness Windows desktop companion: tray icon (whale), desktop shortcut, login auto-start that opens exactly the client you chose (desktop app window or web browser), and one-click desktop/web switching (local Edge/Chrome app mode).
Self-owned desktop shell for DeepSeek Harness (DSH): a small plugin that opens the web surface in a native Tauri window — real macOS fullscreen (decorated window), single instance, window-close-shuts-harness lifecycle. No tray, no bundled Chromium.
QQ Official Bot adapter for DeepSeek Harness (migrated from openclaw-qqbot)
DSH web plugin: create a one-click Windows desktop shortcut (launcher .cmd + auto browser open + taskbar pin hint) for DeepSeek Harness or any local server, right from the Settings page.
OpenAI Codex (ChatGPT Plus/Pro) provider for DeepSeek Harness, with Web UI and CLI device-code OAuth login.
OpenAI-completions-compatible adapter for custom gateways (vLLM / LM Studio / self-hosted proxies): always role:"system", enable_thinking / chat_template_kwargs / reasoning_effort driven by the llm-pi-ai model config, and a </think> split on the receive side — no pi-ai behavior guessing.