プラグイン

Chat-first draw.io plugin: the agent creates, reads and edits native .drawio files, an interactive draw.io viewer renders them in the transcript, and the diagrams live under the workspace's diagrams/ directory. No draw.io account, no npx MCP server, no local editor install.

Bridges the official COROS MCP server into DeepSeek Harness with a self-contained local stdio-to-HTTP bridge and on-device OAuth login, exposing 34 COROS tools on the current endpoint for sleep, HRV, resting heart rate, recovery, training load, activities and FIT files, plus workout and training-plan CRUD and scheduling.

Big Fat Fish clean — unofficial graphical panel for the local Mole CLI on macOS: live system health, cleanup previews with explicit confirmation, disk analysis, app inventory, maintenance, cleanup history, and a guided Mole CLI install.

Adds a Tavily search provider to the DSH web search seam, keyless by default so it works without an account or API key.

Wires the local cc-computer-use MCP server into DSH with one process per root session, giving each session its own Xephyr sandbox screen and private AT-SPI bus (requires the cc-computer-use MCP server).

File dedupe and unify toolkit, shown as "小视频整理" in the DSH sidebar: exact matches by head/middle/tail sampling or full byte compare, near-duplicate MP4s grouped by duration and a configurable first-N-seconds frame window (default 10s, 1–5 fps, default 2; longer videos are compared on their first N seconds too), batch rename or move by extension, UUID, file size or video length, empty-folder cleanup under the selected sources, and MOV/AVI/WMV detected by file header and transcoded to MP4 keeping resolution, frame rate and audio parameters. Requires ffmpeg and ffprobe on PATH. Concurrent workers per task with a shared setting; one task at a time; per-group help dialogs; local-only API. Every task that moves files has one-click restore: per-group/target _dupfc_restore.json records, deleted files skipped and counted, files outside the record renamed to UUID and moved into resetFail; orphaned ffmpeg processes left by a crashed host are swept on plugin load.

Stars the GitHub repositories behind installed DSH plugins from a tab beside Installed on the Plugins page and a control on each plugin detail page; resolves a token from the environment, a saved credential record or the gh CLI, and asks once before the first star because starring is public.

MCP and skill manager as a full sidebar page: add, edit, enable, disable and delete MCP servers, expand a server to its tools and open one for its parameter list, translate skill and tool summaries with a model you pick, and count tool and skill invocations. Also registers an mcp_service tool so the agent can install MCP servers in conversation after an approval prompt.

Private relationship notebook: scores each day of chat from message-level evidence and tracks how the relationship develops over time. Chats, profiles and drafts stay in your browser's local IndexedDB and the plugin runs no server of its own, but using model replies, profiling or automatic analysis sends the relevant chat context to the model service configured in DSH.

"Oh, you're finally here. I've been waiting for you." A plugin that brings Elysia into DeepSeek Harness. Once installed, she remembers you: what you said last time, what you care about, so every new conversation picks up where it left off. Memory lives on your own disk, mirrored across two drives, shared by every session and editable by hand. She speaks the way she actually speaks, measured from 1,696 officially confirmed voice lines rather than a hand-written character sheet. One-click voice setup: the plugin downloads the 200 MB voice weights, places them correctly and fills in the paths for you (the 27 GB engine itself must be installed separately, and needs an NVIDIA GPU). She also drops a WeChat-style bubble with her avatar after finishing a task so you can tap to listen, sends stickers when the mood fits, and learns the runbooks you give her. Memory pack, voice, and stickers are all off by default.

Ship OpenTelemetry GenAI observability and multi-device cloud state sync from one DSH bundle. It reuses and replaces the standalone @loongsuite/dsh-plugin (traces + metrics for sessions, agent loop, LLM and tool lifecycle; do not install both — disable or remove the upstream plugin first), and adds a WebSocket client that pushes full session event logs (header + snapshotEvents) to a relay you run so different local DSH instances share state. On connect it pulls and rebuilds the other devices' sessions automatically (cross-device resume). The sync client itself has zero third-party runtime dependencies; a small Node.js relay is included in the repo. The local /cloud-sync/* HTTP routes (status, sessions, config, resync, delete) have no origin or loopback check, and the bundled relay accepts unauthenticated connections unless AUTH_TOKEN is set — keep the DSH web port off untrusted networks and run the relay with a token.

Adds a /clear command that clears one conversation's model-visible context while keeping the session itself.

Splits a session into turns and incrementally clusters them into named topics and stages, with per-topic token / tool / time stats, a workflow view, a turn timeline, a relation graph and per-turn AI analysis.

DSH plugin: bring your ChatGPT subscription into DeepSeek Harness — Codex models, GPT Image, quota and recovery checks, plus a per-plugin scoped HTTP(S) proxy. Community alpha fork of dsh-codex-connect, published on the alpha dist-tag; it registers the same openai-codex provider, so it cannot be installed alongside the upstream plugin; the image, quota and proxy features are opt-in and off by default.

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.

Adds ChatGPT sign-in, Codex subscription quota (5-hour and weekly windows) and Codex-backed web search to DeepSeek Harness; installing it switches the profile's web search provider to openai-codex.

Runs the local llama.cpp Qwen3.5-4B from a Settings page — Windows only, NVIDIA GPU (8 GB VRAM handles the 4B Q4) — with start, stop and restart, launch parameters, deployment paths with a presence check, provider-route and compaction-policy sync in the profile patch, a loopback request-capture proxy, and a local_worker tool that hands translation, OCR and video-frame chores to the local model by file path. Needs llama-server.exe (verified on b12069) and Qwen3.5-4B GGUF + mmproj files on disk; video frames also need ffmpeg.

Adds two-level tabs (provider and model family) to the composer model picker for large catalogs, with configurable family rules.

Model quality testing for DeepSeek Harness Desktop with a question bank, manual and scheduled checks, saved responses, human reviews, and model attribution. Install it into the desktop profile: the shared Go backend is started only by the desktop app, and a web profile can only view an already-running desktop backend.

ModelScope connector: OpenAI-compatible chat models, a Models-page provider entry, and a composer magicube balance while a ModelScope model is selected.

AIGC creation inside DeepSeek Harness: generate images, video, audio and 3D with open and closed-source models — parameter schemas, cost estimates, approval gates and a task panel.

Lists RouterAI prompt and completion prices for the profile models, highlighting the active one. Requires a model provider whose config key is `routerai`; without it the table shows no data.

Routes model and reasoning effort per turn: a cheap classifier reads each request and picks one of five tiers, a supervisor model re-routes on tool errors and checkpoints, and upgrading to the stronger model waits for confirmation.

macOS-only frosted-glass skin for the DSH NEXT / DSH Desktop conversation pane and third-party sidebars: native vibrancy material, independent switches per level, an optional wallpaper backdrop with adjustable frost, and a Settings section; background material only, no layout changes.