Plugins

Compaction fix for NInfer-hosted qwen3.8-27b gateways: disables thinking on dsh context-compaction and session-title calls so xhigh reasoning stops burning the whole output budget; the same idea applies to other launch methods, this package targets NInfer only.

Unified LLM request/response inspector: reasoning-effort tuning, external-think export, traffic & bundle analysis.

Results-only conversation toggle for the Web GUI: hides thinking and tool-call process rows, keeps one live status line while the agent works, restores working animations under prefers-reduced-motion, zh-CN/en localized. Works in the stock composer and inside dsh-split-screen panes; verified on DSH 0.1.1 and 0.1.5.

Two gates for outbound AI text. (1) Guard: before any text goes out it blocks the shapes anyone could write — padding verbs plus abstract nouns ("optimised the code structure"), hollow adjectives, universal filler, template openers, git commit/PR/review boilerplate, comments that narrate the code. Covers chat, comments, email, commit messages, PR descriptions, reviews, issues, docs and code comments. Passes with OK, blocks with one line naming what to change. (2) Catch memes: when someone drops a slang term the model does not guess or look it up — it already knows. A built-in 6792-entry meme wiki is queried before the model starts thinking, and the entry usage field ("how to answer this") goes straight into context. Matches only exact title/slug/alias hits, never fuzzy, so ordinary chat never triggers it. Bundled SQLite by node:sqlite, zero dependencies, 1.6ms per query. Settings page: guard on/off, strict mode, prompt section toggle, meme wiki toggle, per-turn auto-lookup, a four-step tone slider (reserved / measured / follow the room / unrestrained) that is behavioural rather than a label, AI-completion switch (off by default, uses the session own model so no API key needed), a paste-and-test box, and three thresholds. No model call for any check.

Adds a squeeze tool: the agent hands a long web page, a batch of pages, a text file or a text PDF (via pdftotext) plus one question to a helper model and gets back only the answer, so the material never enters the chat. A settings page picks up to three helper models from the DSH model list, each with a nickname, a fixed thinking level and a "when to use" note, with optional fallback and size limits taken from each model's context window.

All-media (video/audio/image) sibling of liustack/modlens: turns local media into structured evidence (timeline / transcript / layout regions) injected into the conversation so text-only models can cite concrete frames, seconds and speakers — Evidence, not imagination. One failover chain per media, recorded hop-by-hop in meta.attempts (a fallback is never silent): host-registered multimodal models first, then a configured OpenAI-compatible endpoint (comma-separated multi-key rotation), then local ffmpeg frame-sampling / a local ASR CLI, degrading to metadata-only as a last resort. Absorbs modlens battle-tested rules — cooldown keyed only on auth/quota failures and softened to chain-tail (never dropped), redact-before-clip error text, magic-byte sniff that never trusts file extensions. Optional vision-bridge registers a media-lens wrapper into the host model selector so a pure-text model gains image/audio/video sight. Registering / authenticating multimodal models and true inline audio-video upload lean on two companion plugins, declared as optional peers (absence auto-degrades, never errors): dsh-thinking-levels fronts the model config / registration surface, dsh-llm-openai-completions takes over the OpenAI-compatible transport that serializes file blocks into video_url / input_audio. Privacy: media goes to remote providers by default; strictLocalOnly keeps content on-machine. Prerequisites not shipped with the package: ffmpeg/ffprobe on PATH (frame sampling / reencode / duration probe; degrades to metadata when missing) and a user-configured local ASR CLI for the audio fallback. Hosts 0.1.7 / 0.2.0 / 0.2.1 (6 rc, exact engines.dsh enum). npm: dsh-media-lens.

Desktop pet for DSH: 兰酱, an angel-winged maid in a transparent always-on-top window who reacts to session states (thinking, working, waiting, done, error), shows task progress and DeepSeek balance, and responds to drags, head pats, pokes and wing touches.

Keeps DSH reasoning in Simplified Chinese and auto-expands every Think row in the Web UI.

Replaces the composer's model and reasoning-effort selector with a two-level menu and six slider skins.

Shows JeV-selected task plans, progress updates and stage summaries in DSH chat, with native translation or lightweight summaries. Requires a JeV key; the Jev service is called externally and may bill. and hierarchical serial/parallel task flows.

One-tap connect Agnes language, image and video models into DSH (off by default), with image recognition and thinking level filled in automatically. After signing in to the Agnes console (re-login uses AGNES_PASSWORD), you can see the quota windows (5-hour requests, weekly requests, daily images, video), account-wide cumulative usage, and a usage pill beside the input box. AgnesCode desktop models can be connected too (Windows only), re-reading the desktop app login when the token expires, with rate-limit hits backing off on their own.

Think tank: a built-in library of 161 classic mental models (decision/strategy/cognition/psychology/communication/learning/system/innovation/execution) — enter one question and several selected models analyze it in parallel via AI, producing a synthesized report with consensus, disagreements, blind spots and an action checklist; supports both automatic analysis and prompt round-trip import.

Circuit breaker for idle agent turns: watches agent/assistant-stream and cancels a turn that produces only reasoning — no text and no tool call — after 45s, after 80000 reasoning characters, or on a degenerate repetition pattern (exact repeat unit, sentence-template repeat, sparse repeat, or 6-gram density). On trip it injects a continue instruction that requires the next action to be a tool call, and ships 11 regression cases plus per-gate tuning.

Auto-expands streaming thinking: always keeps the latest thinking block expanded, manual actions take priority, only the last block expands when opening a session, and nothing remains after uninstall.

Reasoning-wait thinking indicator for the DeepSeek Harness web composer

Explicit model routing for Main, Subagents, web_search, and image generation, plus a catalog-driven Composer and Plan Review picker with Model, Effort, Context, Fast, and Thinking variants.

A Taffy mood mascot for the DeepSeek Harness Web UI: the corner GIF reacts live to agent state — approval pending / approved / rejected / begging, thinking, tool runs, interrupts, ask_user_question waiting, compaction, long-task fatigue, idle-ignored approval, idle / sleeping — with SSE-pushed state, drag positioning, middle-click console, tool-card mirror console, per-mood locking, persisted appearance settings and per-day bucket statistics.

LongCat-2.0 LLM adapter: 1M context, binary thinking mode, tool calling, and streaming over the OpenAI-compatible endpoint.

Customize the visible DSH Web thinking status with lifecycle-safe CSS.

Adds the Miaoshou (Meituan CatPaw) desktop client as a DeepSeek Harness model provider. The 12 models mirror the ones the Miaoshou app picker shows, each carrying its credit-cost multiplier; thinking levels are translated into the upstream modelParams; a read-only panel reports available credits and the current plan. Sign-in is a QR link, so the desktop client does not have to be running.

Always-on-top status pill for DSH on Windows: shows each session's turn state (thinking with a running timer, waiting for approval, done, failed) above every window, and while DSH is in the background it draws a bubble under the pill to approve a tool request (allow once / reject) or answer the agent's questions without switching back to the app; it declines to intercept whenever DSH is in front or the overlay is hidden, so the official cards still handle those. Sessions can be listed one per row (row count configurable) or folded into a +N badge by priority, the pill collapses to a small dot after a state settles and hides itself while a fullscreen app is in front, clicking brings DSH to the foreground and dragging moves it, and it reuses the Electron build already shipped with DSH Desktop so nothing extra is downloaded. A settings card on the plugin page adjusts scale, mini size, position, layout, row count, hold times, effects, bubble toggles and per-state visibility.

Puts Grok Bot's own face in the DSH sidebar session list: each row's status dot becomes a living GrokBot avatar whose six motions read as idle, working, waiting, blocked, thinking and done — color marks the project, motion marks the state.