Plugins

PK 2-10 models on the same prompt, files, and tools in DeepSeek Harness.

Adds Token Market as an OpenAI-compatible model provider with DeepSeek V4 Pro, Kimi K3, and GLM-5.3.

Persistent plan-mode toggle button in the DSH Web composer: a two-state control (Plan / Plan ×) that turns plan mode on and off, replacing the shipped show-only-when-active badge.

Pressing Enter with an empty composer steers the first queued message into the running turn while the agent is busy.

Session-header button that previews and downloads the exact request frame DeepSeek Harness sends to the model: the full system prompt, the declared tools array, and the wire message array (system/user/assistant/tool-result) rebuilt from the session log via the same folds the agent loop reads.

Integrated memory and context management plugin: four-layer memory system (L2-L4), context compression/trimming/freezing, prompt orchestration, input box UI buttons, and enterprise security with role-based access control.

Real-time DeepSeek API balance (official endpoint) and today-spend estimate (this DSH instance session-log usage x per-model price table) for the DSH web: composer dock strip, always-on floating badge, full settings dashboard (7-day trend, per-model cost, editable price table and polling, history rescan). The API key never leaves the host.

Installable DeepSeek Harness Web UI for replaying a completed session step

Minimal English-only voice input for the DeepSeek Harness Web UI: a mic button in the composer that transcribes speech into the draft via the browser's native SpeechRecognition API. No dependencies, no subprocess, no network calls beyond whatever the brow

A per-workspace HindSight switch for the DeepSeek Harness composer: turn memory off for one workspace without uninstalling it anywhere else.

DeepSeek Harness dashboard: per-session statistics (tokens, timing, tool and skill calls) as a dedicated Statistics view tab. Compatible with dsh v0.1.2-alpha.1 ~ rc.1 (see README).

Auto-send every DSH turn result to WeChat: listens for task completion, extracts the final assistant message and pushes it raw via the WeChat iLink bot (real delivery check, never fake success). Requires the wechat-acp daemon, a separately installed process you log into WeChat with.

Decides whether Python code is original or a derivative work: reports the longest identical expression span against an upstream tree, an exemption rationale per finding, and a risk list. Pure stdlib `ast` analysis — offline, no model calls.

Auto-switch DeepSeek models: routes every request to the vision model when images are present, to the pro model for complex tasks, and to the fast model otherwise; includes a switch_model tool for manual overrides.

Model switching, model capability configuration, fixed compaction models, subagent routing, and input history enhancement.

Native macOS desktop widgets for DSH: manage any number of widget instances — share quote tickers (5-provider fallback, trading-phase-aware throttling) and notes memos — each an independent NSPanel+WKWebView window spawned detached from the host, dragged anywhere, with a type-aware right-click menu. Instances are edited as cards in a three-view settings page (list / type picker / detail), and a market_quote model tool serves standardized quotes to any session. A Type Studio flow seeds a creation-mode session with an issue-template prompt for authoring new widget types.

Multi-agent frugality defense plugin: read-ledger dedup metrics, compaction-immune rules, completion gate, cheap-model review lane.

MaxCompute (ODPS) data analysis plugin for DeepSeek Harness: metadata browsing, cost-gated SQL execution, background jobs and result export.

Windows-only tray monitor for DeepSeek Harness: deploys and manages the standalone DshTaskWatcher tray app - four-state tray icon (gray=disconnected, green=idle, amber=running, red=attention), a per-task panel with live duration/tps/tokens, and completion & outcome notifications. Start/stop switch in Settings > Plugins; the watcher is an independent process that survives DSH restarts.

Auto-injects user-chosen skills into every DSH session, per prompt or once at session start, with a settings page, a composer indicator, and log-deduplicated session-start stamps; verified against DSH 0.1.3 source.

Translate text-based PDFs via DeepSeek or OpenAI-compatible APIs while preserving layout, fonts, images and links, with editable output.

Scoped persistent memory for DSH: memory_save/search/replace/remove/list across global, user, project and failure scopes, DSH storage-domain persistence with secret and prompt-injection scanning, session-scope search, retention cleanup, and optional session-start injection; standing context via memory_pin/memory_pins/memory_unpin.

Auto-renames DSH sessions with LLM-summarized titles: runs after the first prompt and on /handoff, or manually via a header button or /autotitle, with configurable title language and max title bytes.

Force the model to write all replies in a chosen language (Simplified Chinese by default) via a system-prompt section, switchable anytime with a tool.