dsh-reconnect
Safe model-request retry for DeepSeek Harness with exponential backoff, model-availability recovery, and configurable unknown-error limits.
308 results
Safe model-request retry for DeepSeek Harness with exponential backoff, model-availability recovery, and configurable unknown-error limits.
Content-free request and stream diagnostics for DeepSeek Harness
Chat mode for the DeepSeek Harness web GUI: a Chat/Work switch above the conversation, and the managed Chat workspace its conversations are kept in, so a question that is not about a project does not need one
DSH 插件:在聊天记录中自动将本地音视频/图片路径渲染为可播放的预览组件。When an assistant message or tool result contains a local media path, the path is replaced inline with a playable <audio>/<video>/<img> element backed by same-origin /api/media-preview/* route.
A DeepSeek Harness Web plugin for inspecting redacted LLM HTTP request and response traces.
DSH web plugin: 模仿 deepseek 网页版界面右侧的问题条 — 右缘一条竖向问题栏,常态收起为窄列小横线,鼠标靠近时向右展开,列出当前会话每一轮的用户提问,点击后平滑滚动定位到对话对应位置。Mimics the deepseek.com right-edge question rail.
Task-complete & question notification sounds for the DeepSeek Harness Web UI. · 任务完成与提问提示音:Agent 完成任务或向你提问时播放提示音。
DSH session intelligence: a persistent right sidebar with first/latest user-request summaries and read-only Git status.
Standard handoff protocol so DSH never restarts itself: a host plugin that files a JSON reset request for an external ops agent (e.g. Hermes) to execute preflight-snapshot → restart → health-check → recover, then delivers the result back to the requesting session.
DeepSeek Harness plugin: fingerprint the provider, model id, sampling, offered tool schemas, and system-prompt prefix of each request, and diff the live card against a pinned one through the modelprint_check tool
Review-first request-dialect preflight for custom DeepSeek Harness providers.
DSH plugin: anonymize workspace-external file paths before model requests with user confirmation
Unofficial ServerChan watchdog for DeepSeek Harness: send a mobile notification when an approval, plan review, or question remains unanswered.
DeepSeek balance & token-cost plugin for DeepSeek Harness (dual-face): query official balances of every DeepSeek provider (multi API key), compute costs (last question / session / today-project / today-all) from token usage with a user-editable price table. Unified modal UI, bilingual (zh/en)
Turn progress rail plugin for the DeepSeek Harness Web GUI: a sticky dot strip over the conversation scrollport with per-turn question summaries, one-click deep history loading, and AI-option summaries
DeepSeek Harness plugin: route an agent request to a vision model (deepseek-v4-flash-vision-exp) when the user message contains an image, and keep the original model otherwise.
DeepSeek Harness (DSH) conversation index plugin: a dot rail on the right edge of the chat — each dot is a user question; hover shows an abbreviation, click jumps to it.
A GitHub-connected visual git panel for DeepSeek Harness — guided flow (branch → commit → push → PR → review → merge), commit graph with PR/CI badges, one-click GitHub login, and structured git tools for the agent.
Persistent, searchable, per-project memory for the DeepSeek Harness: decisions, rules, and sprint context in a queryable DuckDB file, rules enforced on every model request, with a full management UI in the Web Client
DSH web plugin: a right-side question navigator for the current session. Lists every user question and scrolls to the matching message on click.
DeepSeek Harness plugin: control, organize, diagnose, design, personalize, and answer product knowledge questions for a Yeelight smart home, powered by the local yeelight-home runtime.
Explicit minimal-solution coding guidance for DeepSeek Harness. Registers six load-on-request Ponytail skills and a localized Settings page.
Native desktop notifications for DeepSeek Harness (dsh): run completion, model questions, and approval requests on macOS, Linux, and Windows. Includes a web settings card consistent with the dsh web UI.
DSH web plugin: play a chime when a turn completes or your input (approval / question / plan review) is needed — foreground or background.