Plugin

Adds per-route context window and compaction threshold fields to the Models settings page and a vendor-aware thinking-effort picker to the composer, writing llm-pi-ai capacities, reasoningEfforts, and a compaction-basic patch row.

Visualize the plugin and model call flow as a 3D office with six fixed workstations, thought bubbles and a delivery sheet.

Three tools for DeepSeek Harness: typeset LaTeX formulas into images, draw mathematical figures from a declarative spec, and convert a formula or SVG into an image ready to embed in a document.

LLM provider plugin for the DeepSeek Harness (dsh): streams Claude through your Claude Code subscription (claude.ai OAuth credentials), no Anthropic API key required.

Multi-pet desktop companion: four MAM status lights watch every active session (approval / running / done-unread / error) and each card click-through to its conversation; four voice groups speak on state changes with duration-timed subtitles; import pets from four sources (folder / zip / Codex dir / Petdex registry) with card-style hot-swap and an integrity guard; five action bindings, drag physics, three scales. The built-in fox Foxbell ships in the package — install and go.

Orchestrates durable multi-agent teams from one DSH session: a captain spawns member subagents, drives a dependency-aware task DAG with quality gates, and coordinates through persistent mailboxes, with a live activity panel.

Reads AI replies aloud as they stream, with voiced waiting phrases while the model thinks. Edge TTS built in, any OpenAI-compatible cloud engine supported.

Resume screening plugin for HR in DeepSeek Harness: ingest up to 100k Word/PDF/Excel resumes into one searchable library with skills / coarse categories / ingest time, auto-ingest & de-dupe new resumes, then screen by plain-language criteria — rule pre-filter in the database plus LLM fine-judge on the shortlisted few, CSV export.

Settings section that reports token usage: totals for input, output, cache read, cache write, reasoning and call count, with breakdowns by model, day, month and project plus a per-day trend chart for the last 30 days; history is rebuilt once on a worker thread and then kept current from live session events.

Three-party collaboration protocol for dsh: Step List state machine, five-tier review chain (Gemini/web/Claude Code/dialog/manual), alternatives adjudication, cross-restart resume with heartbeat, and capability persistence.

Opt-in mobile access to DeepSeek Harness Web through QR pairing and a managed FRP tunnel.

Automatic model-tier routing for DeepSeek Harness: one user instruction enters, one tier decision comes out — complex intent is planned on the strong tier and implemented on the cheap tier, simple intent stays cheap throughout, with a deterministic guard and a TTL escalation fallback.

Declare whether a configured third-party model accepts image (multimodal) input; writes the modality into the owning provider settings and verifies it through runtime model resolution.

Sakura-afternoon pastel skin for the DSH Web UI: light and dark sakura token palettes, a click-through falling-petal overlay with a soft vignette, a header petal toggle and a General-settings master switch.

DSH web plugin: open and edit .docx files in the better-sidebar editor through SuperDoc (superdoc.dev) — a browser-native DOCX editor. Assets are self-hosted from this package's node_modules; edited documents export back to disk atomically.

Minimal English-only voice input for the DeepSeek Harness Web UI: a mic button in the composer that transcribes speech into the draft via the browser's native SpeechRecognition API. No dependencies, no subprocess, no network calls beyond whatever the brow

dsh-plugin: encrypted backup / restore / migration of the dsh home config through a private GitHub repo

Unofficial host-side ServerChan alerts for DeepSeek Harness: sends a mobile reminder when an approval, plan review, or question stays unanswered past a configurable threshold, even with the browser closed.

Browser automation for DeepSeek Harness: agent-driven real-browser control (managed Chrome with an isolated persistent profile, or attach to your daily Chrome over CDP) with snapshot, click, fill and page-reading tools; records manual operation flows and replays them with variable slots to create test data on demand; ships 27 browser_* tools plus a Settings panel.

Native macOS desktop widgets for DSH: manage any number of widget instances — share quote tickers (5-provider fallback, trading-phase-aware throttling) and notes memos — each an independent NSPanel+WKWebView window spawned detached from the host, dragged anywhere, with a type-aware right-click menu. Instances are edited as cards in a three-view settings page (list / type picker / detail), and a market_quote model tool serves standardized quotes to any session. A Type Studio flow seeds a creation-mode session with an issue-template prompt for authoring new widget types.

DSH web plugin: a 收藏夹 (favorites) feature in the conversation header — star the current session, list every favorited session, and open one from an anchored panel. Persisted through the `favorites` settings namespace.

Poll Feishu (Lark) group and direct chats through lark-cli, extract candidate tasks with a review workflow, and two-way sync them with the task-board plugin, with a floating browser panel and one-click device-flow authorization.

Drop-in web_search provider that follows the session's routed model: a runtime-editable priority chain with three wire dialects (Anthropic web search, OpenRouter online, Zhipu bigmodel), per-route endpoints and credentials, a Settings UI panel, local JSONL call audit, and built-in official fallback.

Scan, view, and safely clean local AI terminal session histories (Claude Code, Codex CLI, WorkBuddy, DSH) from a settings card — quarantine-first with SHA-256 manifests, one-click restore, audit log, and phrase-confirmed hard delete.