Plugins

Bring the WorkBuddy accounts signed in on this machine into DSH, managing CN and international accounts separately. Supports automatic failover between accounts, one-click refresh of model catalogs and credits, the CN daily check-in, and a togglable OpenAI-compatible endpoint.

Use plugins from ecosystems such as Claude Code and Codex in DSH: add Git plugin sources and marketplaces, upload plugins and skills, add MCP servers, automatically discover and toggle local skills, and manage plugins, skills and MCP in one place, including plugin-provided subagents, Hooks and LSP.

Enhances data visualization in DeepSeek Harness by automatically selecting chart types from user input and generating images for trends, comparisons, proportions, relationships, and flows.

Drives a real Chrome through a local MCP server so the agent can search Google/Bing/Sogou, read any page including JS-rendered and login-only content, automate click/scroll/type/key-press/select in a background window, and run platform search/detail/comment/download for Zhihu, Xiaohongshu, Douyin, Bilibili, X and Reddit; resumable comments, SHA-256-verified downloads, cookies never returned, localhost only.

Task panel for the DSH Web sidebar: publish tasks, auto-distill related sessions, and drive plan/develop/review stages with subagents plus in-panel acceptance.

Adds an approval-mode toggle next to the permission selector: default approval keeps per-call confirmation, bypass approval auto-approves every tool call while staying in Workspace Write.

Cross-session memory plugin for DSH (v0.6.1, npm @asherliner/dsh-memory-connect): auto-extraction, SQLite FTS5 indexing, RRF semantic recall injected into the system prompt every turn, scheduled maintenance, LLM-powered consolidation via ctx.llm, context explosion prevention, and global Soul identity via ~/.dsh/soul.md. v0.5.0 added temporal context graph (valid_from/valid_until/supersedes, append-only corrections via reviseMemory), trust model (recalled history injected as untrusted reference, current instruction always wins), turn-end summarization. v0.6.0 added local semantic embeddings (BAAI/bge-small-zh-v1.5 via embed_server.py, cosine + RRF fusion with FTS5, no cloud calls, ~232 msgs/sec on CPU). v0.6.1 fixed ~ expansion in path config. Zero-config install.

Automatically and intelligently routes each delegated task to the best configured model using aliases, capability tags, and usage guidance—keeping routine work fast and affordable while sending complex work to stronger specialists. Configure your model team in the Web UI or guided setup, run specialists in parallel with reliable result joining, and see the chosen model at a glance in subagent views and inline in Better Sidebar's Tasks tree.

Per-model proxy routing for LLM requests: selected models go through a proxy (Clash etc.) while the rest stay direct, with automatic retry on transport errors, 429 and 5xx, configured live from the settings page.

Real-time DeepSeek account balance badge in the sidebar footer with odometer roll animation, multi-currency support, balance trend chart, per-key usage statistics with historical backfill, and GitHub version auto-check.

AI-driven drawio: four agent tools (validate, render to SVG/PNG, edit, template) plus a sidebar live board — zero-dependency mxfile-to-SVG preview, embedded diagrams.net editor with save bridge, and auto-open whenever the agent draws.

Wakes the main agent after any DSH restart so interrupted work resumes automatically — adds a restart tool and an optional canary that validates the boot and aborts with an alert before a broken restart.

Vision for text-only DeepSeek via Doubao Web by default (zero-cost, no API key — drives your logged-in Chrome through a Windows CDP bridge), with Antigravity IDE quota (flash/pro) or Gemini fallback; auto detail escalation, vision evidence memory with compaction rehydration, content-hash cache, and a bilingual client panel.

API key rotation pool for your LLM providers: auto-discovers providers from settings, per-provider multi-key round-robin, failover on 401/403/429, and cooldown recovery.

OpenCodeGo quota monitor for the DSH Web GUI: a breathing indicator at the input's bottom-right (green/yellow/red by remaining share), a liquid-glass panel with rolling/weekly/monthly usage windows and reset times, auto-refreshing every 30 s; API key read from DSH credentials.

Conversation node navigation rail for the DSH Web UI: one dash per user message on the right edge, reading-position tracking, hover preview card, click-to-jump with highlight line, sliding window, and auto-hide.

Auto-migrate Codex desktop-pet skins into DSH: an animated corner pet that mirrors agent state (thinking, tool use, awaiting approval, failed, done), with per-session frosted-glass dialogs and a full settings panel.

Restart DSH: configurable restart method (Node native / legacy PowerShell), post-restart continue prompt, optional watchdog auto-relaunch.

Web appearance suite with a 7-theme gallery (each a full 122-token --dsw-* mapping) plus one-click switching persisted in settings, Material You wallpaper color extraction that generates a whole dynamic palette, and a global corner-radius control. Themes ride on the official look system as override layers instead of fighting it, so the official light/dark/auto row stays valid. Frozen archive: keep the pinned version, as the author no longer adapts it to newer DSH releases.

Registers two MCP servers that drive the installed AutoCAD and SolidWorks desktop apps over COM, plus two on-demand skills. Windows only, Python 3.11+; the pinned upstream servers install themselves on first load, and only the sides whose CAD application is registered get a row. Each server advertises two meta-tools instead of its full catalog (a discovery shim here for SolidWorks, upstream DISCOVERY_MODE=search for AutoCAD), taking the schemas loaded before the first request from 23,404 to 256 tokens and from 83,732 to 365 respectively.

Discovers other DSH instances on the local network by UDP multicast or port scan, lists them in a sidebar panel together with the capabilities each one reports, and dispatches tasks between machines in both directions over the LAN or over an automatically downloaded userspace Tailscale node.

Auto-classify DeepSeek Harness sessions via Jev decision models or Agent tools, displayed as title emoji prefixes (wip 🚧, blocked ⏸️, review 👀, done ✅, plan 💡, bug 🐛). Features seamless lifecycle recovery, optional Jev inference (needs a Jev API key), model tool maintenance, and a visual settings panel.

Old conversation is folded into short summaries by a side model before the window fills (on-disk ledger survives restarts and re-anchors after host compaction), so long sessions stop getting more expensive every turn — one real full-day session: 6 folds replaced 209 old messages and the live context stayed in the tens of thousands of tokens. The session model picks the model per subtask across six pools (economy/mechanical/main/hard/vision/review); or pick Auto and a small intent classifier routes each message itself, switching only past a confidence threshold to keep prompt caches warm. A three-line watermark ends in an automatic four-step handover (fork backup, handover doc, compact, wake) without losing in-flight subagents; /ctx-pause, /ctx-resume and /ctx-handover keep it manual when wanted. Optional Agent Teams mode adds teammates and a shared task board; /vision gives text-only models eyes via the vision pool; tiered verification sends bigger diffs to a tester plus an independent reviewer. Settings UI, READMEs and language preferences cover the same 11 languages. Bundles three skills: read-only MySQL/Redis verification, commit-message generation, and a requirements/design document convention.

One-tap connect SenseNova Token Plan language and image models into DSH (off by default), with image recognition and thinking level filled in automatically. On a rate limit it backs off and retries on its own instead of misreading it as an exhausted quota. After signing in to the SenseNova console you can see the usage pill, credit pools and per-model consumption. Xiaohuanxiong can also be signed in by QR code, bringing in its models and their consumption multipliers, and taking over DSH web search.