dsh-profile-switch: Interactive Plugin for Switching Named Model Configurations
`dsh-profile-switch` is a DSH plugin designed to uniformly manage default models, thinking levels, and subagent settings through named configurations. It supports both TUI and Web interfaces, enabling fast workspace configuration switching and maintenance via the `/profile-switch` and `/profile-cfg` commands. It simplifies workspace context switching into a single command, eliminating tedious model adjustments. The plugin features headless-mode fault tolerance, supports independent subagent override settings, and depends on the `dsh-subagent-registry` plugin. Configuration data is stored in local JSON files, with self-healing capability in case of file corruption. It integrates deeply with `dsh-tui-pi` to ensure a real-time TUI switching experience.
Read MoreDSH Pin-Sessions: Pin Sidebar Sessions for Quick Access
`fallow5/dsh-pin-sessions` is a DeepSeek Harness (DSH) plugin designed to address navigation difficulties caused by an overly long session list in the sidebar. It allows users to pin active sessions to the top of the sidebar and hide them from the workspace, while centrally managing hidden sessions through an archive panel that supports bulk operations, pagination, and grouping by workspace. The plugin depends on `dsh-session-context-menu` to extend the right-click context menu, adopts a standard host-client separation architecture, and optimizes the user experience for frequently switching between multiple workflows.
Read Moredsh-web-ding: Implementing an Agent Turn-End Notification Sound in the Browser
`dsh-web-ding` is a DSH Web client plugin that synthesizes notification sounds using the Web Audio API to provide idle-state prompts when an Agent turn ends. The host side listens for idle transitions in `agent/status` and emits signals, while the browser side mirrors this namespace. It supports three modes: frontend-only question triggers, Agent turn-end triggers, and Windows 11-style Toast notifications. Messages are stored locally, and volume, frequency, and decay duration are configurable. When using the plugin, be aware of the browser’s autoplay policy and background-tab suspension mechanisms; the AudioContext must be warmed up on first use.
Read Moredsh-balance-pie: DeepSeek Harness Balance Monitoring and Historical Tracking Plugin
dsh-balance-pie is a visual balance-monitoring plugin for DeepSeek Harness (DSH), designed to help developers precisely track API call costs. It displays the initial balance, actual consumption, and remaining quota in real time through a floating pie chart in the bottom-right corner, and supports dragging, minimizing, and adaptive sidebar layouts. It also features actual-consumption calculation, automatic top-up detection, a GitHub-style historical heatmap, and tabular record keeping, with all data persisted locally. Developers can install it via the command line or adjust parameters such as the query interval and price multiplier through the Profile configuration file to achieve fine-grained cost management.
Read MoreDeepSeek Harness Plugin: dsh-gpt56-ptc
The DeepSeek Harness plugin dsh-gpt56-ptc is maintained by community developer Exception-H and is designed to provide cache-stable pure PTC/Code Mode sessions for the GPT-5.6 series of models. The plugin uses static constraints to prevent scope creep and requires DSH 0.1.0-rc.8 and Node.js 20 or higher to operate. After installation, the profile must be restarted, and the model must be manually specified in the "GPT-5.6 PTC" preset to ensure the model remains unchanged during a session. This is an experimental version that does not configure base capabilities, and the preset remains available even after the plugin is uninstalled.
Read Moredsh-find-plugins: Retrieving DSH Plugins by Trust and Relevance
`dsh-find-plugins` is a DSH plugin developed by Eray114514, designed to address the difficulty of finding high-quality plugins within the DSH ecosystem. The plugin aggregates seven data sources, including dsh.so and GitHub topics, and provides the `find_dsh_plugins` tool. It uses a multidimensional ranking mechanism based on "relevance × trust × freshness × popularity" to accurately match plugins, avoiding the noise caused by name matching alone or star-count rankings. The plugin uses read-only operations and in-memory caching, makes no model calls, and requires Node.js >= 22 along with the required DSH version for installation. It supports multi-keyword search.
Read Moredsh-free-games: The Game Cabinet for DeepSeek Harness Web
`dsh-free-games` is a standalone community plugin in the DeepSeek Harness ecosystem that integrates game entry points from the Spawnd.gg platform at the bottom of the Web sidebar. It requires no additional configuration, supports search by name and sorting by release date, name, or most recently played games, and allows click-to-play. The plugin strictly limits its data source to Spawnd and records recent play history using localStorage, making it suitable for quickly adding lightweight entertainment features to Web interfaces.
Read Moredsh-todo-enhancer: Convert LLM-Returned Decision Lists into Editable Checkbox Tables
dsh-todo-enhancer is a Host/Client dual-half-zone plugin designed to eliminate the cumbersome process in DeepSeek Harness (DSH) where users must manually copy and paste LLM-generated lists for decision-making. By recognizing the `<!-- dsh-todo:start -->` and `<!-- dsh-todo:end -->` markers, it renders decision content into an editable checkbox table. It supports inline editing, adding or deleting rows, selecting all, sending directly, and populating a draft, and ensures the panel is displayed only after streaming output completes, thereby improving the interaction experience.
Read Moredsh-matrix-skin: Black-first Matrix Hacker Skin for DeepSeek Harness (DSH) Web
`dsh-matrix-skin` is an independent community plugin licensed under the MIT License, designed to provide DeepSeek Harness Web (DSH) with an immersive, hacker-aesthetic interface. This skin adopts a black-first design, featuring dynamic Matrix rain, terminal-style treatment for Think lines, and full sidebar shell coverage. Core features include a motion-reduction setting, local privacy handling (no network requests or telemetry), and access to reasoning text. Users must ensure they are using DSH `web` profile version 0.1.5-rc.1 or a compatible version. After installing it via the `dsh plugin` command and refreshing, the skin is enabled, delivering dynamic visual effects generated in real time based on the current text.
Read More@tirol/dsh-language-set: Separately Control Response and Thinking Language via System Prompt Injection
Developers using DeepSeek Harness (DSH) often face inconsistent AI language output. The `@tirol/dsh-language-set` plugin independently controls the language of AI responses and thinking through system prompt injection, eliminating the need to manually modify prompt text. It supports Chinese and English configurations, provides commands to view status and injected content, and is suitable for scenarios requiring strict language control. Installing the plugin requires a service restart; verify the source code and permissions before use.
Read Moredsh-pentest-bugtrace: BugTraceAI Penetration Testing Mode Plugin
dsh-pentest-bugtrace is a penetration testing plugin designed for DeepSeek Harness (DSH), aimed at simplifying the process of building security testing agents. The plugin encapsulates BugTraceAI’s MCP tools and workflows into a standard plugin, with built-in penetration testing roles, authorization rules, and preset runbooks. Users only need to configure environment variables and run a specific command to activate the agent for authorized scanning, vulnerability validation, and report generation. It supports multiple MCP tools to drive the BugTraceAI engine and includes a fail-open policy to handle backend service unavailability, greatly reducing the configuration barrier for penetration testing agents.
Read Moredsh-win-toolkit: Windows Native Capability Plugin
The `dsh-win-toolkit` plugin provided by DeepSeek Harness (DSH) uses secure PowerShell scripts to provide agents with core Windows capabilities, including clipboard read/write, system notifications, hosts file reading, and network diagnostics. The toolkit supports operations such as clipboard writing and notifications (permissions required); the remaining functions are read-only. Inputs are transmitted via Base64 encoding to prevent injection. Installation requires executing a specific command and restarting the service, and requires a Windows 10/11, PowerShell 5.1+, and Node.js 22+ environment.
Read Moredsh-netassist: DeepSeek Harness Network and Proxy Diagnostic Assistant
dsh-netassist is a DSH toolkit plugin maintained by Edge-Echo, designed to provide agents with read-only network diagnostics capabilities. It uses PowerShell on Windows environments to help users troubleshoot GitHub connectivity, system proxy status, and port liveness issues through DNS resolution, TCP port probing, HTTPS status code checks, and Hosts file scanning. The tool supports end-to-end diagnostics for any host, and all operations are read-only. After installation, agents can trigger checks using natural language. Note that it only supports Windows, and end-to-end diagnostics such as `net_doctor` can take a relatively long time.
Read Moredsh-gacha-calendar: DeepSeek Harness Gacha Game Schedule Sidebar Plugin
The DeepSeek Harness sidebar plugin “Gacha Game Calendar” is designed to let you view the current banners and event schedules of major anime-style games with a single click. Core features include a floating sidebar panel displaying schedules and countdowns, automatic online refresh (with manual forced refresh and interval refresh), independent multi-source data with automatic fallback, and custom entry management. The plugin can fetch data from official announcements, wikis, or third-party sites, and includes parser self-check functionality. Users can install it via a command. It is mainly used to aggregate and display publicly available schedule information and does not store resources.
Read MoreIntroduction to DeepSeek Harness Local Model Plugin
`dsh-plugin-local-model` is a plugin for DeepSeek Harness designed to simplify local model inference. It uses a dedicated settings page to automatically launch llama.cpp upon the first conversation message after a model has been configured, and it automatically unloads after 5 minutes of inactivity to release resources. The plugin also supports one-click switching between parameter presets, ensuring offline operation and file security. Users must prepare their own models and tools; for installation, it is recommended to refer to the source code. This tool is suitable for developers who need local deployment, privacy protection, and frequent parameter switching.
Read Moredsh-session-unarchive
DSH 2026-10-01 added a session archiving feature, but it is hidden by default in the Web interface, preventing users from viewing or restoring archived sessions. The `dsh-session-unarchive` plugin resolves this as a pure Cordis plugin without requiring modifications to the DSH installation directory. It adds an “Archived” button to the sidebar, displays a list of archived sessions, supports title filtering and one-click restoration, and offers both Chinese and English interfaces. Users must install it via dependency injection; it no longer relies on file patching.
Read Moredsh-self-improvement
The DeepSeek Harness Agent plugin provides agents with a maintainable long-term memory pathway through tiered and weighted mechanisms. It categorizes memory into five types, supports retrieval based on importance and recency, and offers governance features such as deduplication, poisoning protection, and score-based eviction to manage capacity constraints. The plugin also supports automatic retrospection, skill accumulation, and cross-session experience reuse, ensuring that agents retain context after restarts or session switches. Users can manage memory using commands such as `/memory` and `/retro`, and can configure memory scope at either the workspace or global level. It is suitable for complex application scenarios that require long-running operation, automatic experience accumulation, and secure governance capabilities.
Read Moredsh-voice-talk: Voice Conversation Plugin for DeepSeek Harness
`dsh-voice-talk` is the voice plugin for DeepSeek Harness, designed for hands-free use. Through full-screen voice interaction, it enables simultaneous speaking and listening, as well as real-time content generation and read-aloud. It supports Chinese and English, streaming interaction, and adjustable speech rate and voice timbre. Before using it, install host version 0.1.0-rc.7 or later and configure an Alibaba Cloud Bailian API Key to enable speech recognition and synthesis. Plugin keys are stored locally and follow the principle of least privilege, making it ideal for efficient use in Chrome or Edge browsers.
Read Moredsh-ego-browser: Site Skill Memory for the DSH Browser Agent
`dsh-ego-browser` is a memory plugin for the DeepSeek Harness browser agent, designed to solve the problem of repeatedly learning site tool signatures and operation notes to improve efficiency. This MIT-licensed open-source plugin provides core features such as `ego_recall`, `ego_run`, and `ego_learn`, supporting tool registration, inherited storage, and environment checks. Installation requires Node.js 20 and `ego lite`, and has been validated in dsh v0.1.7-rc.2.
Read MoreDeepSeek Harness Plugin: lightroom-py (Mount a Python Library to Control Lightroom Classic)
The DeepSeek Harness plugin `dsh-lightroom-py` mounts `lightroom-py` as a native tool, enabling agents to interact directly with Adobe Lightroom Classic. It covers a broad range of features, including catalog management, photo operations, metadata editing, developing tools, mask creation, and collection management, and supports automation via the CLI. The project is licensed under the MIT license and is unofficial. It currently offers better support for macOS; on Windows, services need to be started manually, and some AI features remain unavailable due to SDK limitations.
Read Moredsh-prime-memory: Hierarchical Distilled Memory Plugin for DeepSeek Harness
The `dsh-prime-memory` plugin for DeepSeek Harness is designed to address memory loss in long conversations by leveraging a hierarchical distillation mechanism (L0-L3) to automatically consolidate and inject historical information. It offers session tier switching, a cost dashboard, tool calls, and a “write-only, no-read” mode, all transparent to users. Installation requires Node ≥ 22.16, and data remains in a local directory after uninstallation, making it well suited for handling complex Agent tasks.
Read MoreIntroducing the DeepSeek Harness Plugin: dsh-perm-gate
`dsh-perm-gate` is the permission gate plugin for DeepSeek Harness, combining approval, classification, and gating into a single package. It uses a fixed priority chain and a fail-closed mechanism to ensure that high-risk operations cannot be bypassed. Core features include command blacklists, session authorization, LLM-based risk-tier adjudication, and adjudication learning. The plugin supports a pure-function rule engine, an audit log stream, and an automatic review mode, resolving cross-plugin version coupling issues while providing a unified audit trail and fine-grained control.
Read MoreIntroduction to the dsh-llm-openai-completions Plugin
`dsh-llm-openai-completions` is an adapter in DeepSeek Harness for connecting to custom gateways, primarily resolving compatibility issues related to role definitions, thinking mode, and visual input. Its core features include enforcing the use of the system role, separating thinking content, supporting image uploads, and adapting to various Qwen model formats. Users can install it via the DSH command line and enable it in their configuration file. Please note that maintenance for this project was discontinued in October 2026, remains at version 0.1.0, and the 0.2.0 version with visual features was never released.
Read Moredsh-context-compression-improved: Context Compression and Code Skeleton Compression Plugin for DeepSeek Harness
The `dsh-context-compression-improved` plugin in DeepSeek Harness addresses context bloat in agent tasks through mechanisms such as pre-compression, aggregate replacement, and history compaction. Its core new feature, the Code Skeleton Gate, substantially reduces oversized source-code tool results by retaining only declarations and key lines, significantly improving inference efficiency. The plugin supports multiple strategies, including Fresh and Aggregate, is compatible only with DeepSeek models, and requires manual enabling of code skeleton compression in the settings.
Read Moredsh-openspec: OpenSpec Workflow Skills and Command Alias Plugin
The `dsh-openspec` plugin is designed to address the integration of DeepSeek Harness (DSH) with the OpenSpec workflow. It registers OpenSpec’s 12 core skills as `openspec-*` specification skills and provides `/opsx-*` command aliases, supporting invocation by the model and users. The plugin depends on Cordis, is easy to install, manages state through a standard skills registry, and requires no additional runtime configuration.
Read More