Plugins

Pixel-art Little Red Riding Hood pet for the DSH session window: transparent and always on top, it walks back and forth along the composer, blinks, looks around, turns its head toward the cursor, sleeps on double-click, and scales from 2x to 7x in a right-click panel.

Whale-girl desktop pet for the DSH Web UI, drawn with hand-written SVG + CSS (no image or audio assets, zero runtime dependencies): swims while a turn runs, blows labeled bubbles on tool calls, taps the glass with an amber pulse when an approval is waiting, jumps with a turn duration/tool-count report on success and droops on failure, naps after dark or when idle. Click to poke, double-click to pet (affinity hearts), drag with remembered position, right-click menu (feed / mute / reset / hide), hover card with level, title, XP bar and per-turn stats. A `whale_say` tool lets the agent make her speak or act as a lightweight visible notification. Read-only over sessions and workspace; only ~/.dsh/whale-girl/state.json is written.

Pre-execution guard: a `tools.guard` predicate decides allow, ask or refuse per permission preset for dangerous shell commands (git hook bypass, catastrophic `rm`, download-and-run, git history rewrites) and for edits to credential paths or to files the turn never read, and hands the model the reason plus a working alternative.

Turn-end quality gate: when the turn edited code or config it walks each file up to its project root, runs that root one check from a fixed whitelist (`pnpm check`, `cargo check`, `ruff check .`, `tsc --noEmit`) and, only if that check really ran and failed, opens a new turn carrying the failure output instead of letting the turn close.

macOS Desktop only: one-click restart of the DeepSeek Harness Electron app. Confirms before interrupting running turns, blocks during plugin install/update, and records restart phases/logs.

Turns what a design, plan or document claims into a model and decides it: UML diagrams rendered, read and reviewed; state machines, protocols and data models checked as models; migrations tested for coverage and rollback symmetry; concurrency claims mined and routed to the checker that can decide them; dependency graphs audited; reports diffed.

Speaks a cue for DeepSeek Harness on Windows: a configurable phrase when an agent turn completes normally, a second phrase when the agent needs you to choose — a question, a plan approval, or a permission request — and a time-of-day greeting when a new conversation starts. A small bubble in the lower-left corner mutes and unmutes every cue with one click, with its own toggle chime. Phrases are synthesized with a Microsoft Edge neural voice (skipping aborted and errored turns, subagent turns, and rounds that a goal is still driving), cached under the DSH home directory and replayed locally, with the desktop SAPI voice as the offline fallback.

Flashes the DeepSeek Harness taskbar button on Windows when an approval, a question or a plan review is waiting, or when a turn finishes.

Turn-finished music reminder: when an agent turn ends, play a ringtone or any local audio file (bundled synthesised telephone ring, custom absolute path, or an in-page upload), with a floating chip to pause, resume or dismiss it and loop-until-paused playback; configured in Settings, defaults overridable per deployment.

Toasts from DeepSeek Harness when a turn finishes, and when it stops for approval, a question or a plan exit. Native on Windows, text and timing configurable.

Flashes the Windows taskbar icon in the DSH desktop app when a turn finishes in the background or an approval or question is waiting for you.

Plays a voice alert and a browser notification when a turn finishes, asks for input, or fails, and can push the same summary to a Feishu bot webhook as a card.

Turns real git history into two Chinese changelog drafts (short for users, long for developers); every entry carries its commit SHA and an evidence-backed verified/unverified mark, and commit subjects are never translated.

Deterministic ticket triage: a rule prefilter classifies clear cases and returns everything else as needs_review instead of guessing.

Expose a running DeepSeek Harness instance over MCP, so an external agent can list, create and continue that instance's own sessions, delegate a turn and collect its result, answer the questions the DSH agent asks, and set how future approval requests resolve.

Injects a very short conventions reminder (absolute INDEX path + version) once per session per version; when the conventions version changes, the next turn re-injects it (chat turns included). Optional hard guard: on operation-like turns, write/edit tools are rejected until the session has read the index (chat turns and shell are not blocked by default). Ships example checklists or points at your own directory; zero dependencies, configurable paths. Solves: conventions are written down but never pulled up when the work starts.

Keeps a DeepSeek Harness run going on its own: continues a turn that ended abnormally, and continues a turn logged as completed while the model tool call was emitted as plain text instead, which leaves the agent silently idle. Revives a model request that stops streaming. A wait on an approval or on a question is logged and notified, never answered.

Arms a watch on a directory and delivers a batched, deduplicated diff when the tree goes quiet, filtering out the writes the session made itself; delivery is per watch, either waking a new turn or queueing for the next step.

Skill-creation console for the Web UI: fetches the official creator skills, hands the requirements to the Huawei Cloud, generic, or REST API creator, then scores the generated skill against a 10-point rubric and returns the itemised deductions to the creator until it reaches 8.5/10.

Shows which skills each turn invoked and installs, updates or removes skills inside the app.

Registers two tools for DeepSeek Harness. `doc_ocr` turns scanned PDFs and images into structured text (per-page text, financial tables with a balance-sheet tie-out check, contract-clause hits) through a local RapidOCR engine or a Zhipu GLM-OCR cloud engine, and `doc_preprocess` turns a directory of raw files into an Obsidian-ready knowledge base through a seven-stage pipeline with backend routing and sha256-based incremental skipping. Numbers that fail the tie-out are reported, not invented. The local engine needs Python 3 with rapidocr-onnxruntime, pypdfium2 and Pillow (pip install -r requirements.txt); the GLM engine needs ZHIPU_API_KEY.

Structured image recognition: local sharp measurements plus a cloud vision model return description, OCR with coordinates, object locations and fine-grained attributes in one call; a fingerprint cache makes repeated images return in milliseconds without spending model quota.

Tencent Yuanbao Search provider for the ctx.web seam: one direct Search API request per query returns portable title, URL, snippet and ISO-8601 date, with credentials resolved per search and no auxiliary model turn. Requires a Tencent Cloud WSA API key.