Plugins

Article Review page in DSH's native right sidebar: turns Markdown, Word, PowerPoint and PDF articles into sentence-addressable text, lets you comment on any sentence, and hands the open review to the agent as a composer draft. Six agent tools let the model read every comment, revise the article (in place for Markdown, as a sibling <name>.revised.md for binary formats), resolve the comments, and record recurring wording preferences into a long-term writing-style memory injected into later prompts. Extraction includes a dependency-free DOCX/PPTX reader; PDF text extraction runs locally through the host's Python + pypdf (or poppler's pdftotext) and is unavailable without one of them. No network calls, no telemetry.

Windows companion window for DSH Desktop: todos, reminders, live session messages, approval handling and quick input. The standalone window reuses the Electron binary at %USERPROFILE%\.dsh\electron\electron.exe (usually installed by dsh-pet) and fails to open when it is missing.

Save any URL (intranet GitLab raw, public GitHub, etc.) to a local file — text or binary — with a text preview and skip-on-exists.

Exposes a DSH Harness Desktop Web GUI to another machine's browser through a Cloudflare quick or named tunnel, an SSH reverse tunnel, or a custom command, and hands out a tokenised link to open it; a settings panel and CLI start, stop and print the link. Needs DSH Desktop 0.2.1-rc.2 or later and a cloudflared binary.

DSH host bundle for the ShyPie privacy-first workflow: extracts attached images and PDF/Office files locally before they enter the session log (requires poppler pdftotext/pdftoppm on PATH and a local OCR engine — Windows OCR on Windows, optionally tesseract; without them attachments are rejected, not forwarded); default filter mode keeps text local (optional local masking); blocks unmaskable media/BLOB; provides local privacy_extract, privacy_ocr, privacy_scan and privacy_rehydrate tools; denies read_image and binary/document reads outside allowPaths; and enforces a fail-closed llm/stream egress guard before DSH dispatches a model request (the default provider allowlist is the local Magpie gateway).

Audio tools for DSH: inspect audio/video files, transcode or extract the audio track from a video, trim and normalize, and transcribe speech with local whisper.cpp or an OpenAI-compatible API. Requires ffmpeg and ffprobe on PATH (not bundled); local transcription also needs a whisper-cli binary and a ggml model.

One-step local model onboarding plus a light harness for DSH. On load it probes loopback and private-network inference endpoints (llama.cpp, Ollama, vLLM, LM Studio), reads each model's context window and output budget from the backend, and registers the provider through settings, so a local model becomes selectable without hand-writing config. It then trims only that provider: masks the tool list through tools.restrict, drops named system-prompt sections and memory contexts, swaps the injected instruction block for a short one while preserving its first 80 and last 96 characters, and truncates tool-schema descriptions without touching structure or required fields. Cloud providers are returned as the same assembly object, untouched. The injected block is re-read from disk on every assembly, so editing the file takes effect on the next request with no reload. The panel in the plugin detail page is a live operations view: it shows whether the takeover is active and by which hook, how many characters the last assembly was trimmed from and to, and which providers are online (probed concurrently, with a real one-shot completion reporting latency and reply text per provider). If a local backend is not running, the panel offers a one-click launch: it probes the port first (never starting a duplicate), then starts the matching server (ollama serve, or a custom command you supply) detached and polls until the endpoint answers, clamping the wait to 3-600 seconds, reaping the child it just spawned if it never answers, and naming the paths it searched when it cannot find the binary. A stop button kills backends this plugin started (SIGTERM, then SIGKILL) and refuses to touch a server somebody else launched. When a model has just been downloaded and no backend exists yet, the panel can also scan the machine for .gguf files with hard caps (300 models, 20000 directory entries, Downloads only one level deep), start llama-server on the chosen one (skipping the .mmproj. vision projection and reusing a server that already answers), register the provider under the model id the backend reports, and run one real completion to prove it works. Before spawning anything it does a fit check that counts only the part of the file the GPU cannot hold: hostBytes = fileSize - VRAM, compared against max(total RAM - 2GB, free RAM). A file far larger than RAM is refused outright; a heavy host-side remainder is refused unless you confirm; the same remainder under mmap is only a warning, because the weights land in reclaimable page cache. A load-mode selector (auto / mmap / none / mlock / mmap+mlock) maps to llama.cpp --load-mode, and the default is whatever the engine picks. From 0.4.2 everything above is derived from the platform instead of assumed: home and data directories come from os.homedir() and the OS conventions (LOCALAPPDATA on Windows, ~/Library/Application Support on macOS, XDG_DATA_HOME elsewhere); backend binaries are looked up in the official install locations of each OS and can be pinned with OLLAMA_BIN, LLAMA_SERVER_BIN, VLLM_BIN or LMSTUDIO_BIN; model discovery walks DSH_LOCAL_MODELS_DIR, ~/models, ~/llama.cpp/models, ~/.cache/llama.cpp and the LM Studio data directory, with the WorkBuddy layout demoted to one candidate among them. VRAM is asked from nvidia-smi, then rocm-smi, then sysctl hw.memsize on Apple Silicon (flagged unified, so the same bytes are not charged twice), and any flip to a unified-memory budget is handled explicitly. Before building a command line the plugin runs llama-server --help once and keeps only the flags that build actually has, so older engines get --n-gpu-layers 99 and no KV quantization while newer ones get auto plus q8_0; anything it cannot probe is passed through rather than dropped. LM Studio is started with lms server start --port N rather than being lumped in with the engines that need a model argument, and vLLM gets its own --model example instead of llama.cpp syntax. Windows, macOS and Linux are all exercised in CI, including a windows-latest and a macos-latest job, because a cross-platform claim that has only ever run on one OS is not a claim. Two of those suites exist specifically to catch this machine's blind spots: one replays the whole suite with process.platform rewritten to each OS, and one deletes every advantage this development box has — no nvidia-smi, no rocm-smi, no sysctl, an empty PATH, a home directory with no models — and requires graceful degradation rather than a crash. That last one caught a real class of bug: several assertions hardcoded 'lms.exe', which cannot pass where the executable has no suffix. Measured on the reference machine: a 16.9GB 27.2B MoE with 14.5GB of VRAM holds a 285-691MB working set and sustains 76 tok/s after a 1000-2000 token warm-up, so file size alone is not a fit signal. From 0.4.3 the scanned models are shown as a table rather than a dropdown (one row per gguf, with its size and any fit warning), so you can see what is on disk before choosing. From 0.4.4 the delete button on the provider table removes the provider entry, not the weights file on disk. That distinction matters: what people actually want to clear out is the dead entry - one pointing at a port that never answered, one registered twice for the same port, one whose model file was deleted long ago - not a 15.8GB gguf. Deleting a config is undone by importing again; deleting a gguf is not. The removal goes through settings.mutate with an unset op on \['providers', id], because update merges layers: writing null there leaves a null provider key that the loader reads as a broken config rather than an absent entry, and only mutate/unset is implemented as Reflect.deleteProperty. The result is read back afterwards, because unset on a key that comes from a lower layer (the profile cordis.patch.yml or a home patch) does not delete it - it restores the inherited value, so the call succeeds while the entry stays. When the entry is still there the panel says so and points at the file to edit instead of reporting success, and a DSH whose settings service predates mutate() is told plainly rather than silently ignored. Deleting takes two clicks and clearing the confirmation is as easy as clicking another row; on success the online and baseURL caches are cleared and the tool surface is recomputed, since the provider list changed. The gguf itself is never touched by this plugin - the scanned-model table is display and selection only.VRAM reporting is driver-dependent. From 0.4.5 model discovery also reads Ollama's own storage: its blobs are named sha256-… with no extension, so extension matching could never list them. Weights are now accepted by extension OR by GGUF magic header (only dot-free files of at least 8MB, capped at 600 header reads per scan; obviously-not-weights extensions like .json/.md/.safetensors/.part are never opened). Ollama models are resolved from their manifests: the model layer is picked by mediaType application/vnd.ollama.image.model (never adapter/mmproj layers), the human-readable name is restored from the manifest path (…/qwen3/8b becomes qwen3:8b, …/latest stays bare), tags sharing one blob merge into a single row, and the metadata kv supplies name, quantization and context length. Adopting an Ollama model does not start a duplicate llama-server over the same weights: it confirms ollama serve answers on 11434 (launching it if not) and registers the provider under the tag the endpoint itself reports — the blob's sha256 name would never match anything, since 11434 lists qwen3:14b, not sha256-a8cc…. Two real-machine-only traps fixed on the way: existsSync('<file>/') returns true on Windows, so manifest-vs-directory branching now goes by parseable content instead of trailing-slash semantics; and the quantization label table had been copied from a stale ggml enum generation, labelling ggml_type 15 as Q5_K_S while ollama show reports Q4_K_M. Same-blob dedup is by (dev,ino), keeping whichever entry has a name. Verified on the reference machine: the scan goes from 1 row to 4 — qwen3:14b (8.64G) and qwen3:8b (4.87G) now appear with Q4_K_M badges — in 4ms; npm run verify exits 0 (launcher 57, portable 35, panel-host green including the linux/darwin simulated-platform runs); the new panel-host suite builds a real miniature Ollama store in tmpdir and points OLLAMA_MODELS at it, so results no longer depend on whether the test machine happens to have Ollama installed. From 0.4.6 the panel shows each model its real name instead of its file name, because some download tooling stores every weight under one generic file name, leaving the real name only inside the GGUF header. The scan now reads each file GGUF KV section - not a fixed-size head buffer, because tokenizer.ggml.tokens (hundreds of thousands of strings on a large MoE model) and the chat template push general.file_type past any window; and GGUF string arrays are interleaved (u64 length plus payload per element), so treating them as a length table walks the cursor off the rails and produces astronomically large reads. Reading and skipping are accounted separately: key names and types are always read, string values over 256KB and string arrays are only walked past without copying, non-string arrays over 64 elements are skipped outright, and a cursor past the file size stops the read, because an out-of-range position means the cursor is wrong rather than that a limit should be raised. The displayed name goes Ollama tag, then general.basename plus size_label, then general.name, then the file name, with the full string kept for the hover text since it usually carries internal build numbers; mmproj files are now identified by general.type as well as by name. A short read (truncated file) no longer feeds half a value into the parser. On the reference machine, reading one file header parses 44 keys in about 16ms, and a whole scan takes about 31ms. From 0.4.7 the smoke test no longer fails the first time and passes the second. The old code gave both request attempts one shared AbortController and one shared 30-second timer, so the total budget covered both rounds: a cold load of a large model burned it in round one, the abort left the signal permanently aborted, and the second round (the Ollama-style reasoning_effort: none retry) could not even be sent - which is exactly why the first click returned This operation was aborted and the second succeeded. Now each attempt gets its own controller and timer, a timeout retries once automatically (retryOnTimeout, on by default) because a first-request timeout almost always means the weights are still being read from disk, and non-timeout errors (refused connection, HTTP error) are not retried at all. Timeouts report themselves: the message says the wait exceeded smokeTimeout, whether a retry already happened, and points at the new smokeTimeout setting (seconds, default 60, clamped to 5-600) - port responding is not the same as tokens coming out, so it is a separate knob from launchTimeout. Unrelated: an example naming one specific model was dropped from the panel copy, since not every machine has that model downloaded, and the explanation now describes the mechanism instead. From 0.4.8 adopting a plain .gguf no longer registers an unrelated Ollama model. The port field used to follow the last click rather than the file: selecting an Ollama entry moved it to 11434, selecting a plain .gguf did not move it back, so a 11434 left over from an earlier click stayed. The server accepted it as given, probed the port, found Ollama answering and borrowed it (the skipped-launch log line about the port already answering is this path), and then the model id - derived from the file name - could not be matched against Ollama's /v1/models, where the fallback was to grab the first entry in the list. Embedding models tend to sort first, so the adopt registered a model the user never selected. Now the port follows the file: a plain .gguf landing on another engine's port (ollama 11434, vLLM 8000, LM Studio 1234) is moved to 8090 and the panel says so. Only ports belonging to a known engine are overridden - a hand-typed 9999 aimed at your own llama-server is left alone. The model-id fallback no longer guesses either: it takes the only entry when there is exactly one (a single-model backend may legitimately report a different id than the file name) and otherwise reports an error listing the ids it did see, rather than registering a wrong provider. From 0.4.9 the panel can no longer vanish. It used to be possible for the whole section to disappear from the plugin detail page while the host stayed up, the HTTP bridge kept answering, and no crash log appeared. The cause was a scope error: a helper was declared inside one component and referenced from another, which is valid JavaScript and passes a syntax check, so it only threw at render time - when the user pressed the scan button and React ran the state updater. The host wraps every slot entry in an error boundary whose fallback is an empty anchor, so the throw was contained silently: the panel was not uninstalled or crashed, it was replaced by an empty div. The helpers now live in the module scope both components share. The client half also gained its own test - a minimal React stand-in loads the real panel module, really calls the components, really runs the state updater at render time, and presses the scan button, so this class of error can no longer reach a release.

Matrix-style binary (0/1) digital rain for the Harness interface: columns spawn only while a task is running, fall off screen when it stops, and leave a blank canvas; frosted-glass token layer with adjustable digit and panel opacity.

In-app browser for DSH: a dedicated Chromium rooted at a persistent profile so a Feishu or SSO login survives restarts, driven over CDP, exposing seven agent tools (status, open, eval, read, feishu, screenshot, tabs) and mirroring the page live in a right-sidebar panel that tool calls open by themselves. Also converts Feishu documents to Markdown by sweeping their virtual scroll. Requires a locally installed Chrome, Edge, Chromium, or a Playwright Chromium build (it never downloads one); set DSH_IAB_CHROME to point at a specific binary.

Folds REA's 207 reverse-engineering tools (Ghidra / Hopper / IDA backends) into 3 commands plus 6 read-only domain entries: about 5 KB of resident tool text instead of 757 KB. Ships no engines; rea-agents, Ghidra 12.1.x and JDK 21 are installed separately.

Snipping tool for the composer: freeze the screen, drag a region, annotate it, and drop the result into the message box. Windows only; the selector is a bundled WinForms binary (rebuilt from the included C# source if missing) and uses the built-in .NET Framework 4.x.

Agent tools that download a Bilibili video, transcribe it with a local whisper.cpp build, then write a structured Markdown note with DSH's own LLM (optionally printing it to PDF). Windows only: missing dependencies are auto-downloaded into the data directory — yt-dlp.exe and the ffmpeg.exe build from GitHub and gyan.dev, the whisper.cpp x64 binary from GitHub, and a ~547 MB model from the hf-mirror mirror; PDF export shells out to a locally installed Edge/Chrome. Two tools: video_to_notes and note_to_pdf. Pure JS, no extra API key.

Remote workspaces for DeepSeek Harness: pick an SSH host in the sidebar, mount a server directory over SFTP as a workspace, then read, write and edit those files with the normal file tools. Remote watch is polled so the file tree refreshes, image/binary previews work through byte-range reads, and remote_exec, remote_search and remote_git give agents the remote shell, search and git queries. An opt-in patch routes the agent's own bash/pwsh commands to the server. Every remote write and command follows the session sandbox mode and approval policy, and hosts are managed by the plugin itself (password, key or SSH agent).

Turns a plugin repository's existing screenshots into a 6-second silent looping hero reel: a VP9 webm that keeps real alpha, a matted gif, and a cover frame, plus a manifest with per-scene holds and sha256s. The timeline is exact by construction (integer holds plus transitions sum to durationMs, and the last scene transitions back into the first so t=0 equals t=duration). Zero runtime dependencies: PNG decode/encode on node:zlib, an SDF card compositor, and ffmpeg as the only external binary, discovered at runtime. Ships a conversation-view panel (pick, reorder, generate, copy the README markdown) and a CLI with a storyboard.json contract.

Mounts the pen.dev desktop app MCP server in a DSH session so the model can read and edit the .pen document open in Pen through the get_app_state, read_skill, execute, get_style and browser tools; finds the server binary by probing the usual install paths, and adds a pen-dev skill plus a listener that turns pen.dev plan-limit failures into a short bilingual notice. Unofficial, and ships no pen.dev binaries.

Pins an always-on-top progress bar while the agent runs a turn — the percentage is estimated from tool calls and filled to 100% when the turn ends — and badges the DeepSeek peak/off-peak pricing window so large token operations ask for confirmation only during peak hours. The bar runs as a separate Electron process and is skipped when no Electron binary is found.

Open STL meshes in the right-Sidebar document preview: browser-side binary/ASCII parsing, a WebGL viewport, and size/volume/surface-area/watertightness statistics.

Registers one `doc_ocr_sign_fix` tool for DeepSeek Harness: on macOS it diagnoses and repairs the condition where the bundled Python cannot load native extensions (onnxruntime, cv2, pypdfium2) because of the hardened runtime and library validation, by injecting the disable-library-validation entitlement and re-signing the interpreter, then verifying the imports. Operations are idempotent and the original binary is backed up first. It never touches the enclosing application bundle.

macOS arm64 only; requires dsh-pet installed (the bundled binary is a closed prebuilt). Desktop switch for the dsh-pet pet: one click to show or hide it, and it can start DeepSeek Harness when that is closed.

Opens STL meshes in the right-Sidebar document preview: binary/ASCII parsing in the browser, a WebGL viewport, and size, volume, surface-area and watertightness statistics.

DSH Cordis computer-use provider for the ZTE Nde desktop. npm install auto-fetches the musl binary built by computer-use-linux-zte-nde (v0.7.7-nde.3) into the package's own bin/ via postinstall; resolveCommand() resolves the cordis command path against import.meta.url so the mount works regardless of dsh web's cwd. Reserves the shared computer-use slot via ctx.computerUse.register and exposes MCP tools under mcp__computer-use-linux-nde-mcp__* (screenshot, list_windows, click, type_text, doctor, etc.). Two notable Nde fixes upstream of this provider: KWin caption+coords fallback for window targeting, and SnapshotCache that keeps AT-SPI single-wake per (app, pid) so Chrome is not repeatedly disturbed.

A mandatory 7-step plan on every request, each step with a binary acceptance criterion, executed through worker subagents and validated by independent memoryless verifier subagents that report PASS or FAIL with evidence.

DSH host plugin that drives the ployenv local runtime version manager from the chat (list/install/use runtimes, ship a ployenv binary, and inject its path list into session shells).

Reverse-engineering workbench: bridges a local IDA Pro into DeepSeek Harness through @deepseek-ai/dsh-mcp-client, exposing the ida-pro-mcp toolset (decompile, xref, rename, comment) as mcp__ida__* tools. Auto-selects idalib headless mode (analyses a binary in-process, no IDA GUI) or GUI mode (connects to the IDA plugin RPC on port 13337). Adds re_check, re_open and re_status; the bundled MCP server is a zero-dependency stdio proxy usable from any MCP client.