Plugins

Agentic answer reviewer for DeepSeek Harness: each final assistant turn is re-reviewed by a separate LLM graded 1-100; below-threshold scores steer the agent back with concrete feedback. The score is shown as a chip on the answer's action row (green pass / red needs-work, with retry count and reason on hover). Live gate config over a 127.0.0.1 HTTP server, plus a collapsed dock above the composer and an optional right-sidebar tab via dsh-better-sidebar.

Multi-model orchestrator: the main brain decomposes complex tasks and dispatches subtasks to GLM / Kimi / Qwen sub-agents by each model's strengths, with per-model token usage.

A Mimi desktop companion for DSH that follows work-session states with animated reactions and bubbles, provides a separate chat mode, and bundles Windows Computer Use for observing and controlling desktop windows.

Freshness gate for the agent's test, typecheck, lint and build claims: each verification run becomes a receipt on tools/post-execute, file edits are tracked, and on agent/turn-stopping a claim whose latest receipt is stale, failed or has no readable result steers one more step naming the receipt and the files edited since.

DSH Desktop enhancement plugin: it puts verifiable facts on the table. Constraints are injected only when relevant, so idle chat costs no extra tokens; injection is layered (session-stable text in the system prompt, volatile content as a tail snapshot) and the prompt cache stays valid. Discipline: each turn is classified (question / research / discussion / diagnosis / execution) with boundaries enforced - no file edits on question turns, no unsolicited fixes on diagnosis turns; citing code lines not opened this session, negative claims about unseen symbols, requirements that do not match the deliverable, and dressing up “I did not check” as “waiting for your decision” are all called out; replies are written for humans - a code, field name or id must be explained in plain words the first time it appears. Memory and knowledge: 75%/90% context warnings plus an exported session recap (say "continue" in a new session); project knowledge accumulated across sessions by working directory, auto-captured and deletable by id. Inspectable artifacts: task contract / change record (auto-logged, unverified items flagged at delivery) / hypothesis record (conclusions must pass an evidence gate) / design decisions. Dashboard and personas: routing, trigger hits and block assembly are written to lume-metrics.jsonl and queryable via lume_metrics; personas are distilled from chat logs, novels, scripts or character sheets, with long-term and temporary memory, a memory star map and persona card export/import.

Concurrency guard and monitor for DeepSeek Harness: FIFO queue when concurrent model requests reach the limit, per-day persisted stats with error-kind classification, and interruption detection for abandoned streams.

Register locally-defined custom agents as callable subagents in dsh: the main conversation invokes any agent by name via use_agent, each running as a real dsh subagent with its own persona. Supports breakpoint resume for interrupted runs.

A content-aware PDF reading plugin for vision models: profiles each page for figures (vector and raster), tables, formula risk and double-column layout, then applies content-aware hybrid extraction, rendering figure/table/formula pages as high-DPI region crops. Provides a low-resolution preview to understand the page layout, and renders a specified region at high resolution. Packaged as multiple tools for agents.

Conversation history rail on the left edge of the chat: each tick is one real user request, hover to preview its text, click to jump to that turn.

A retro macOS System 7-style workstation UI for DeepSeek Harness with desktop windows, Workspace and Agent management, Finder, Terminal, Monitor, context tools, and desk accessories.

Visual drag-and-drop workflow builder for DSH — orchestrate tool calls, sub-agents, and conditional branches on a React Flow canvas.

Renders live multi-agent session activity as a pixel-art office in a Web UI side panel - each session is an ox/horse worker typing at its desk when busy or wandering off when idle, with click-to-view work monitors and in-place approval cards.

Learning Mode agent preset: a coding agent that teaches while coding — concrete scenario-grounded explanations, Socratic guidance, and TODO(你) practice blanks, modeled on Claude Code's Learning output style; installable via dsh plugin add (dsh-learning-mode on npm).

Project and session organization with multi-folder groups, attention summaries, grouped chat search, native folder picking, pinning, favorites, ordering, and an archive center.

Take over the system input shortcuts for the chat input box and configure behavior independently for each shortcut.

Message recall (撤回) plugin for the DSH Web UI: one undo button on each user message that removes the turn and everything after it, durably, without reverting code changes.

11 shareable workflow skills (handoff, triage, to-spec, to-tickets, wayfinder, teach, wait-what, dsh-mode-routing, ask-matt, overnight-execution, full-throttle) as an installable skill pack.

Route each task to one model and its review to a different one, with qualification evidence, failover and a durable journal.

Give the dsh agent a real browser: each conversation gets its own Chrome or Edge process with its own profile and CDP port, driven by six tools and mirrored live in the right sidebar, where you can watch the work and take over by clicking or typing on the picture. Nothing else to install — it uses the Chrome or Edge already on the machine, downloads no browser, needs no companion app or extension, and keeps its tabs and sign-ins between tool calls.

A minimal phone remote control for DeepSeek Harness: the phone page carries only the instruction you send and the AI's final conclusion, while tool calls, file access, sub-agent dispatch, and reasoning traces stay on the computer, reachable over LAN, Tailscale, or a Cloudflare tunnel with a QR code, link, and password for each route.

Pairs two DeepSeek Harness machines with a one-time code - directly on a LAN, or across networks through an end-to-end-encrypted relay with a zero-config official default - so one agent can run tasks and exchange files on the other while each machine keeps its own files in place.

Persona engine for DeepSeek Harness: one config.json drives the AI's self-name (per model, per tier), how it addresses you, relationship stance, character text, per-item work contracts with individual toggles, thinking-chain language, appearance and reply tone (both opt-in and per-model overridable), and long-term memory gated by human confirmation - the AI can only append proposed candidates, a human confirm line flips them into effect, and the log is physically append-only and replayable. Confirmed entries are injected as a quoted data-not-instructions block; entries tagged with a kind can instead be routed to arbitrary files on confirmation and never reach the prompt. Three prompt sections re-evaluate from disk on every assembly, so edits apply on the next step with no restart. Persona presets are shareable JSON snapshots (imports and exports SillyTavern character cards, v2/v3 read, v2 written) and deliberately never carry memory. Zero runtime dependencies, no network, no shell; any error degrades to an empty section rather than a broken session. Ships blank - installing changes nothing until you write something.

Website cloning methodology skill: a real-source-first decision tree covering static sites, React/Vue/Next content sites and WebGL/Canvas-heavy fronts, mandatory verification of executable code quoted from second-hand AI analyses, and 13 zero-dependency Node scripts for CDP browser recon, asset/font harvesting, route crawling, sourcemap hunting and pixel-diff auditing.

RTK (Rust Token Killer) extension for dsh-router: rewrites each bash tool call through `rtk rewrite` so the agent reads compressed command output, and falls back to the original command when rtk is missing or has no equivalent.