Plugins

Delete archived DSH sessions with reference-aware attachment garbage collection, as one reversible transaction.

MCP lazy-loading, dynamic tool-loading, and Tool Router plugin for DeepSeek Harness with fail-open schema progressive disclosure: hide compatible MCP tool schemas until routed to reduce context bloat and save tokens, while explicit lazy connections stay w

Smart network proxy manager for DeepSeek Harness — fixes desktop Transport Failed and TLS interception problems

Ollama Cloud provider for DeepSeek Harness with native and OpenAI-compatible transports.

Native continue button and remappable shortcut for interrupted DSH conversations.

Pinned-sessions block for the DeepSeek Harness sidebar: a collapsible pinned panel above the Settings row, hover pin buttons on session rows and a session-header toggle. Pins live in the plugin's own settings namespace and are migrated once from dsh-sessi

DSH model provider for Cohere — OpenAI-compatible wire by default, Cohere's native v2 lane for citations/RAG/thinking-budget. Direct wire adapter, no proxy.

Sticker and emoji reactions for the DeepSeek Harness Web/Desktop chat: a composer sticker picker, agent stickers that render as images while the transcript keeps the pack tag, emoji reactions under the user bubble, and an emoji-rain effect.

Context Transparency Manager: the model context as a visible, editable, effectiveness-scored surface (self-contained dsh bundle plugin)

Arabic-first actionable Windows notifications for DeepSeek Harness (dsh): toast buttons that continue, stop, or open the app — and approve or reject the agent's permission requests — without leaving your seat. Bilingual ar/en with RTL.

DeepSeek Harness plugin: saved shortcuts (name + target agent + text) with a workspace/session/agent tree picker, a sidebar menu that delivers a shortcut's text to its target without switching sessions, and per-message capture from user bubbles

DeepSeek Harness plugin: VS Code-compatible multi-root workspaces. A .code-workspace file is the single source of truth — its `folders` become writable roots of the session, exactly as in VS Code / Trae / Cursor.

Registers im_say, a tool that sends each call as its own IM message through dsh-im's proactive delivery service, so a single reply can arrive as several bubbles instead of one block; im_targets and im_target_add list and create the delivery targets it sends to.

Session-scoped delivery receipts for local artifacts: compare current file hashes with a recorded baseline, keep agent claims separate from explicit user acceptance, flag changed or unavailable files, and export a Markdown receipt through a tool, slash command, and conversation panel.

Dictation that ends up in the repository instead of the chat log: hold the button (or import a WAV), and the plugin archives voice-memos/<id>/ with audio.wav, a rendered waveform.png, memo.md (frontmatter, per-segment timestamps, ready-to-paste quote blocks) and meta.json. Audio is resampled to 16 kHz mono in the browser before it is sent. Transcription stays local — the harness speechToText service with silence-split windows, or a command worker such as faster-whisper — and when no engine is available the memo is still filed with an explicit status and reason rather than a fabricated transcript. Ships a conversation-view 口述 tab with click-to-seek playback and per-segment quote copy, three agent tools (voice_memo_save / voice_memo_list / voice_memo_read) and HTTP routes. Zero runtime dependencies, 52 node:test cases, real screenshots in docs/.

Prove that a local image edit left every pixel outside its mask untouched. Original, generated and mask PNGs go in; a greppable `MASK_PIXEL_LOCK=PASS|FAIL|EMPTY` verdict, a difference heatmap, per-channel statistics, a bounding box and the largest changed pixels come out, plus a self-contained HTML report and `report.json` with a SHA-256 file hash, a canonical protected-pixel digest and a proof digest binding both to the policy. Supports edit and lock mask polarities, a mask threshold, tolerance and alpha-ignoring; PNG color types 0/2/3/4/6, bit depths 1-16, all five filters and Adam7. Zero dependencies, local only — no model call, no network, no upload.

Fills in the vision capability that DSH's "fetch models" step drops, so image-capable models stop being declared text-only, and marks them in the model name.

Archives DSH conversation output into SiYuan notes: a single siyuan_archive call files content under `<archive path>/<year-month>/<type>/<date>_<time>_<topic>_<content>` with a metadata table and auto-maintains an index document at the notebook root; 6 siyuan_* host tools (list notebooks, full-text search, read document, archive, delete block, remove document) and a Web settings page for base URL with connection test, API token in the host credential store, archive target and title template, plus notebook/document-level read/write/delete permissions with default deny and child levels able to only tighten.

This project aims to bridge the Blender MCP Add-on, so that models on DSH can conveniently connect to the current user's working Blender to assist creation, while meeting basic security requirements.

Local ComfyUI image generation for DeepSeek Harness — discovers or starts the ComfyUI installed on the machine, generates a text-to-image result from a chat request, and attaches it to the conversation.

Auditable token-cost ledger for DeepSeek Harness: wraps the llm/stream waterfall and records every call with provider, model, input/output/cache-hit/cache-write tokens, usage source and integer nano-CNY cost from an explicit price table. Aggregates by hour, day and session; exports JSON or CSV with a no-write dry-run preview; keeps a manually entered official platform bill in a parallel column with actual-minus-estimate diff. Unknown provider/model routes stay explicitly unpriced. Zero runtime dependencies; no balance query, notifications or animations.

Adds a writing console above the composer with four style tiers (T1-T4) and five rigor tiers (L1-L5), selectable by hand or left to automatic choice; keeps project terminology and banned phrasing in .mem/config.md with a read-only human section; and provides a whale_polish tool that rewrites expression without changing facts.

Per-workspace memory library with manual per-chat opt-in, editable rules, three sharing modes, and continuous forgetting; cards stay readable Markdown.

Restart initiated by the agent itself, not by a button. The agent calls one tool, then the plugin takes over: it waits until every agent and background job on the host is idle, runs a visible ten-second countdown that a click, a scroll, or a key press cancels, then restarts DSH and wakes the original session so the agent carries on from its last sentence. No button, no manual entry: the tool is the only way in. Works on the official desktop app and the official web client; third-party shells are not supported out of the box. Windows only.