dsh-precedent
Evidence-backed working memory for DeepSeek Harness: a cited ledger of what already worked in this workspace, built from the session log you already have
630 results
Evidence-backed working memory for DeepSeek Harness: a cited ledger of what already worked in this workspace, built from the session log you already have
Voice practice mode for the DSH web app: bilingual (中文 / English) conversation output plus read-aloud (TTS) and speech input (STT). Adds a 语音交流 dialog-mode toggle to the composer.
DeepSeek Harness (dsh) plugin for machine-wide token usage statistics — a Settings page with totals, a contribution heatmap, streaks, and a per-model breakdown. Reads session logs from every dsh home; no cost estimates.
A DeepSeek Harness (dsh) security plugin: blocks agents from reading or writing sensitive files (.env, credentials, key material), masks leaked secret-shaped values in tool results, keeps an audit journal, and exposes safe inspection tools that never print raw values.
A DeepSeek Harness (DSH) plugin that reads PDF files of any size (no 64KB limit): full Unicode text extraction (Chinese/English) via pdfjs-dist, with automatic OCR for scanned or image-heavy pages (Windows WinRT OCR zh-Hans-CN + en-US, or optional tesseract.js).
Codex-style message-bar navigation rail for the DSH dialog: a vertical rail of short horizontal lines (one per user message) at the sidebar's right outer edge, vertically centered; hover highlights and floats the sent content (first 30 字, 15 字 per line); click jumps to the message.
DeepSeek Harness plugin: let the agent read, fetch and parse Zhihu (answers, columns, comments).
DSH tool for archiving and reading WeChat Official Account articles via 2100laike WeChat Archive (wx.2100laike.com)
Automatic model routing for DeepSeek Harness: a virtual `smart` model classifies every request by difficulty (hard / normal / easy) and by vision need, then delegates to the models you already configured in Settings → Models. 内置智能路由:三级难度 + 视觉自动路由,模型从「设置-模型」已配置模型中选取,虚拟 smart 模型零配置接入。
DSH_HOME single-instance lock plugin: aborts startup loudly when another dsh server already uses the same data directory, preventing concurrent session-log writes (the 'seq gap in committed region' corruption class)
Give a DeepSeek Harness agent its own email inbox, with inbound mail bound to per-thread sessions.
DSH plugin guard: on every plugin load checks profile bundle lists for duplicate loader entries (aggregate bundle + sub-plugins listed together → 'duplicate loader entry id' boot crash) and runs site-level health checks — known conflict pairs, vendor @deepseek-ai tree integrity vs lockfile, plugin @deepseek-ai junction integrity — guarding the 2026-08-18 and 2026-08-21 incidents. Read-only diagnostics.
Hash-anchored read/replace/undo tools for DSH. Every line of a file carries a unique 3-char content hash; edits target hashes, never line numbers, so stale anchors are caught before they touch the file. Port of pi-hashline-edit-pro for the Pi coding agent.
Vision for text-only DeepSeek Harness agents: delegate image reading to a one-shot subagent on a configurable vision route (MiniMax / Kimi / any OpenAI-compatible provider), keeping image bytes and the vision model's context out of the main session.
DSH web plugin: switch the current session's agent preset (mode) at any time, even after the conversation has already started - the stock UI refuses that. Provides a session-header dropdown plus a fenced /preset-switch/api route.
Extend the DeepSeek Harness (DSH) web chat drag-and-drop to accept every file type: dropping any file in the composer saves it into the current session's workspace so the agent can read and use it.
DSH web 插件:折叠思考过程直接输出回答内容;在会话输入框上方增加会话活动状态条,展示 Think / Edit / Bash / Read,可展开与收起。 / DSH web plugin: folds the thinking process so answers render directly, and adds a session-activity status bar above the composer showing Think / Edit / Bash / Read, expandable and collapsible.
DSH LAN file share: share the workspace over the local network — phones/tablets/other computers browse, download and upload files via a browser, no DSH login needed. Token gate, path boundary, read-only mode, rate limit. Zero-dependency (node:http).
DeepSeek Harness plugin: OpenCode Go plan usage readout docked under the composer, aligned with the input bar width
Verification toolkit for DeepSeek Harness agents: evidence-based claim checking against workspace files with line citations, config file validation (JSON/YAML), HTTP URL status checks, npm package checks and GitHub repo submission-readiness checks — complementary to dsh-plugin-gate
让 DSH 直接操控你的浏览器,无需视觉能力:零 spawn 连接你已开启调试端口的 Chrome/Edge,截图、像素断言、执行页面 JS。Drive your already-running Chrome/Edge over CDP — screenshots, pixel assertions, page JS — no vision model, no per-use approval, zero child processes.
China-network bootstrap for DeepSeek Harness (dsh): connectivity probes to npm/GitHub/HuggingFace/mirrors, local proxy detection, mirror recommendations, and a generated PowerShell/bash bootstrap script. Read-only by default; apply is explicit. CLI + agent-callable cn_boot tool.
Quality scorecard for DeepSeek Harness (dsh) plugins: 0-100 score with grade and per-component breakdown (manifest, peer resolvability, dist-tag health, dead ranges, freshness, docs, dsh-tools peer compat). Zero runtime dependencies, read-only. CLI + agent-callable quality_score tool.
Remote multi-server GPU status board for DeepSeek Harness Web: read-only nvidia-smi over SSH with ControlMaster-muxed connections, a host-side cached snapshot, per-server selection, and gpu_overview / gpu_refresh agent tools.