Plugins

Reads a video or audio file into metadata, sampled stills, subtitles and a multilingual speech transcript with speaker labels, captures the live desktop as frames or a recording, and installs its ffmpeg/whisper.cpp/sherpa-onnx toolchain on first use with per-file sha256 checks; all processing stays local. Automatic toolchain install and desktop capture currently support Windows x64 only, so elsewhere ffmpeg and model paths must be set manually.

Drives the DSH right-sidebar built-in browser from the agent side with 38 tools (open, navigate, read page text and interactive elements, accessibility tree, click, fill, select, check, key press, scroll, resize the browser frame by dragging its corner, with the page inside reflowing responsively so a narrowed frame shows the mobile layout, screenshot, full-page screenshot, PDF, markdown export, scrape to JSON, download, cookies, per-step operation records with replay and action whitelisting), serves the built-in web_search from a self-hosted SearXNG over an SSH tunnel with Tavily and a real-browser fallback, and registers a fetch provider that retries blocked GitHub URLs through measured mirrors and sniffs charset from the response header so GBK and GB2312 pages decode correctly.

Loopback relay as DSH's process-wide proxy route: web_fetch works under TUN fake-ip; a dead proxy gets same-request direct fallback, and a probe switches back.

Tavily-backed web search and page fetch for DeepSeek Harness: a Search provider, a Fetch provider with Tavily Extract, Firecrawl and direct HTTP modes, and a settings card. Needs a Tavily API key; Firecrawl is optional.

Connects a ComfyUI server to the agent through five draw_* tools (status, workflows, models, generate, reverse) and a sidebar panel. It ships no workflows and assumes nothing about your models: every API-format JSON you drop into the workflow directory is inspected to find where prompt, negative, width, height, steps, cfg, seed and LoRA live, and a sidecar file can override what was detected. LoRA injection splices into any graph. All paths are configuration, editable from the panel and saved to a settings file, so it works against a remote or already-running ComfyUI. Headless auto-start and idle shutdown are opt-in and off by default.

Computer use for DeepSeek Harness: js and js_reset tools that read windows, click, type, and screenshot macOS apps through the locally installed ChatGPT app. Apple Silicon macOS only; requires the ChatGPT desktop app.

Extracts deadlines, deposit, validity, ceiling, bid composition and copy counts from a tender document, and reports the items it did not find rather than omitting them.

Converts tables in .docx, .xlsx, .xlsm, .csv and .tsv files into the table JSON consumed by the rule checkers, and reports when it could not rather than guessing.

Checks a site safety-qualification ledger for training hours, the four required record elements, and special-operation certificate validity against the dates the person worked.

Checks a payroll sheet against published social-insurance, housing-fund and individual-income-tax parameters, with pass, pass_with_gaps, review and fail outcomes and parameters that carry their own review date.

Checks an invoice ledger for duplicates keyed on code + number, arithmetic consistency, payer mismatch, date anomalies and template placeholders, and emits an anomaly list plus a workbook.

Joins two spreadsheets on a declared key and compares declared fields within declared tolerances, producing a Markdown report, a JSON result and a diff workbook. It reports differences; it does not adjudicate them.

Checks a project document archive for missing required document types, stale revisions and empty files, using folder and file names rather than document contents.

Checks a canteen food-retention ledger for the four required record elements, the 125 g minimum, the 48-hour cold-storage window and template placeholders left in a cell.

Gives the agent 14 tools to create and edit DingTalk Yida apps (apps, forms, pages, schema, data) through the OpenYida CLI, plus a right-sidebar workbench that opens them in the built-in browser.

Optional Decision API tool discovery and search-source prefetch, with per-session controls and ordinary-discovery fallback.

Scores each assistant answer on five dimensions plus topic coverage, shows evidence, and turns the weakest dimension into copy-ready prompt advice. Observational only; local by default, with optional Jev review.

Production pack for ComfyUI inside DeepSeek Harness: submit API-format workflows to a local or remote server, batch-generate and fetch outputs, and diagnose bad images with measured parameter ranges and a failure-mode decision tree (black frames, duplicated subjects, color fringes, OOM).

Run Blender headless from DeepSeek Harness: execute bpy scripts, render scenes and export glTF assets, with modeling cookbook, 5.x pitfall list and asset-pipeline playbooks bundled as skills so the agent self-checks geometry before handing assets back.

Drives Proteus 8/9 interactive simulation from the agent through eleven proteus_* tools: connect, simulation start/stop/status, probe readings, a project save that checks the file hash and mtime actually changed, and layout suggestions computed locally; on machines that register the Labcenter.InteractiveSimulation COM server it also places devices from a whitelist and queries and connects pins, while without COM it falls back to a keyboard-driven UI transport covering the simulation lifecycle and saving. Requires Windows with a local Proteus 8/9 installation.

Drives a local MaaEnd (MXU) client over its loopback HTTP API: 10 maend_* agent tools, a sidebar dashboard panel, and startup-mute / success-gated close-game steps in the task chain. Requires a local MaaEnd install with its MXU service running.

Park whole tool packs out of the session's tool surface and mount them back with one command. Cuts the tool table from 99,789 to 48,144 characters (51.8%) without uninstalling plugins or restarting. Reversible per group, per agent.

Port of the V2 memory pipeline from OpenAI Codex to DeepSeek Harness: a background agent extracts each idle session, consolidation folds the extractions into one memory_summary.md behind a git baseline diff, new sessions get that summary injected, and cited memory flows back as usage counts. A Settings page chooses the model, each stage's reasoning effort, and the idle, age, concurrency and timeout limits.

Cross-session development worklog for the agent: worklog_add appends one line (optional tag) to today's file, worklog_list reads a day back or lists the days on record. Entries are plain JSON under $DSH_HOME/worklog, so the next session can pick up where the last one stopped.