Plugins

Evidence-bound negative-knowledge ledger: persists disproven paths (command_failed, file_missing) with their outcome and precondition evidence, then warns or blocks repeat attempts until the evidence changes.

Archived Chats settings page for DeepSeek Harness: browse, search, unarchive, and delete archived sessions grouped by workspace — deletion takes effect immediately, no restart needed, even for sessions still resident in the background.

Per-subagent LLM provider/model selection with role templates.

LLM provider that runs the locally installed Claude Code CLI as the model backend, so requests go through an existing Claude subscription instead of a metered API key; native tool calls are bridged over MCP.

Bottom-right quota capsule that auto-discovers every configured provider (DeepSeek, OpenRouter, SiliconFlow, GLM, one-api/new-api, coding plans) and shows balance or rolling 5-hour and weekly usage, with per-provider visibility, alert thresholds and proxy support.

Configurable DSH usage statistics: a sidebar text/card widget plus a detail dialog with per-provider and per-model breakdowns, cache-hit rates, token trends, and latency details (TTFT and duration).

Load Wallpaper Engine .mpkg files as the DSH web background: embedded video, time-of-day switching, unified frosted blur, local wallpaper library & rotation.

Client-side writing panel for the Web UI: project volumes and stats, corpus library, full-text search, evolution version-chain diffs, and an SVG thread graph, shown only in writing-preset sessions.

Task board for the dsh web GUI: a sidebar entry plus a multi-column kanban view with local persistence.

Runs open-mcp-apps inside DSH: every MCP app gets its own sidebar container with a separate workspace, session, and App mode, an agent status strip under the app, and inline app rendering in ordinary chats.

A Claude Code-style reasoning-effort slider for the DSH Web UI: stepless drag, snap-on-release, a WebGL fire trail, and real, working thinking-effort control for any custom third-party model or provider.

让 DeepSeek Harness 的第三方 API 支持低、中、高等推理强度，并可为每次子 Agent 调用选择模型｜Add Low, Medium, High, and other reasoning levels to third-party APIs, with model selection for each subagent call

Per-model capability declaration for DeepSeek Harness: reasoning-effort levels (wire spellings) + request modalities (vision) for OpenAI-compatible providers. Settings section, zero runtime harness deps, no YAML.

provider-agnostic AIGC HTTP 桥 + 无限画布 + ffmpeg 后处理，13 个工具含画布连边/reroll/媒体编辑 | Provider-agnostic AIGC HTTP bridge + infinite canvas + ffmpeg post-processing; 13 tools incl. canvas linking/reroll/media-edit

Bridges session images to configurable vision providers and returns text-only analysis to eligible DeepSeek Harness model routes.

Automatic provider/model fallback chains for DeepSeek Harness agents when LLM requests keep failing (retry exhausted, auth, quota, rate limit)

Official-grade GitHub CI integration: a composite action.yml, a polling PR review bot with idempotent inline comments and a status-check gate, plus PR/issues tools with every write gated by human approval.

User-side undo/redo for DeepSeek Harness (dsh): the /undo command rolls the transcript back to before the last message and /redo restores it.

Common Word Box (常用词盒子) — DeepSeek Harness web client plugin: a persistent slide-up word/phrase panel next to the composer input, with global/current-project buckets and one-click insert. Works in every workspace.

Draggable floating card showing the balance/quota of your most recently used providers (DeepSeek/Moonshot/Kimi For Coding): auto-discovery, color-coded tiers, live refresh.

The model switch for DSH: point it at any OpenAI-compatible endpoint, with curated free/cheap DeepSeek provider presets and automatic fallback when a free tier rate-limits you.

Two-level model circuit breaker with failover: trip a model or a whole provider after repeated request failures and route the next request to a configured fallback.

Evidence-first operating kit: six read-only tools (capability catalog, staged workflow plans, packaged skill reader, bounded local memory search, repository release audit, release checklist) and five packaged skills - plans and audits, never remote writes.

Extends dsh-subagent-tools with a per-call cwd for subagents, shipped with the two in-process provider patches it requires.