dsh-gpu
GPU-aware execution layer for DeepSeek Harness: gpu_status / gpu_exec / gpu_run_bg tools, per-step GPU context injection, automatic CUDA_VISIBLE_DEVICES card selection
734 results
GPU-aware execution layer for DeepSeek Harness: gpu_status / gpu_exec / gpu_run_bg tools, per-step GPU context injection, automatic CUDA_VISIBLE_DEVICES card selection
Midscene-based AI UI automation for DeepSeek Harness: one ctx.midscene seam, Android and Web (Puppeteer) providers, and the android_ui / web_ui model tools
Kimi WebBridge for DeepSeek Harness: drive the user's real browser (navigate, click, fill, snapshot, screenshot, evaluate, network, upload, PDF) through the local Kimi WebBridge daemon
Complete workspace grouping manager for the DSH web sidebar: 分类 → 项目 → 会话 three-level tree, manual group create/rename/delete, drag-and-drop grouping & ordering with insertion indicator, rule-based auto-classification, tree search. Sidecar YAML + runtime overlay, zero core-storage intrusion.
Sync opencode-go models from https://models.opencode.ai/api.json into llm-pi-ai provider opencode-go-autosync (one-click create/update in settings).
Windows Computer Use for DeepSeek Harness: window-bound screenshots, robust OCR, verified clicks, pure-OCR mode, pluggable vision models.
Archive manager (browse / unarchive / auto-rules) plus a Codex-style left tick rail for jumping between conversation turns. · 归档管理(浏览、取消归档、自动归档规则)与左侧对话刻度栏跳转。
Recursive Self-Improvement (RSI) engine for DeepSeek Harness: a bounded perceive → integrate → verify → repair → propose loop that keeps a downstream desktop app in sync with the fast-moving upstream harness. AI for AI — DSH maintains DSH.
Codex-style right sidebar for DSH Web with automatic output collection, responsive previews, and session-aware tabs
Hindsight memory manager for DeepSeek Harness: settings UI (intranet/extranet routes) with automatic check & install of the official @vectorize-io/hindsight-coding-agents DSH adapter
Canvas preview panel for DeepSeek Harness — live-preview HTML artifacts, auto-sync, save-as, PNG/JPG/SVG export. DSH 画板预览插件:实时预览、另存为、PNG/JPG/SVG 导出。
DeepSeek Harness 自动审批插件:在 workspace-write 沙箱之上增加 Auto 档——确定性规则放行/拒绝,模糊操作由 LLM 裁决,危险操作转人工。保留工作区沙箱边界,不放宽为 full-access。
DSH 插件 UI 调试神器:让 AI 在真实 Chrome(Playwright)中自动看界面、点按钮、拖组件,一键安装零配置
Gitee/GitHub AI 员工:issue 里 @ 机器人自动开发并提 PR;v1.2 新增代码安全扫描——配置仓库地址后按内置/自定义提示词扫描漏洞,去重后提交 issue。An issue-driven AI developer for DeepSeek Harness (Gitee & GitHub) with optional static security scanning (dedup + issue reporting).
Kimi WebBridge host plugin for DeepSeek Harness: model tools that drive the user's REAL browser (navigate, snapshot, click, fill, evaluate, screenshot, network inspection, tab management) through the local kimi-webbridge daemon
dsh-polyglot — the model switch for DSH: one generic OpenAI-compatible ctx.llm adapter, curated free/cheap DeepSeek provider presets, and automatic provider fallback when a free tier rate-limits you
DSH 微信自动化插件:任务进度微信推送、消息监听、微信⇄DSH 双向桥(设置页开关+配置)。基于 wxauto4。
DeepSeek Harness plugin: auto-repairs an invalid messages array (orphaned tool_calls / tool messages) to stop 400 INVALID_REQUEST session lock-ups.
DSH plugin for controlling mobile devices (ADB / iOS simulator / WDA)
Start a DeepSeek Harness conversation without selecting a workspace: auto-provisions a default workspace and adds a "Chat without a workspace" option to the workspace picker
DSH Web client plugin: adjustable chat column width and alignment, a numbered question history with keyboard navigation and auto-load of older questions, plus markdown outline rails on both the conversation and the file preview.
让 DeepSeek Harness 自动重试商汤 429/QUOTA 等瞬时网关故障直到任务完成——启动时给所有 llm-pi-ai provider 热注入加固版 retryPolicy,运行中新增的 provider 也会自动补上,再也不用手动发「继续」。
DeepSeek Harness plugin: auto-detect secrets pasted into the composer, store them in the official credentials seam, and send [secret:REF] placeholders to the model instead of the value.
BrainAgent — brain-inspired plugin for DeepSeek Harness (dsh): a pipeline of heuristic filters, statistical memory stores and context injections; reward-ledger + UCB1 bandit learning loop (RL-lite)