DeepSeek Harness Plugin: dsh-web-plugin-manager
DSH Harness (DSH) uses a plugin-based architecture that allows its toolchain to be enhanced through extension modules. This article introduces a plugin named `dsh-web-plugin-manager`, maintained by healing1, with its repository hosted on GitHub. Currently, detailed information about the plugin's specific features, installation commands, and usage examples is unavailable. Users should note that the plugin operates with the permissions of the current DSH process, and license information is missing. It is recommended that users directly visit the GitHub repository to review the latest documentation and source code.
Read Moredsh-task-console: The Floating Task Console for DeepSeek Harness
dsh-task-console is a plugin for DeepSeek Harness (DSH) designed to consolidate scattered tasks, subagents, and session data into a unified floating task console. It features a flip-based interface and supports real-time data monitoring, card interactions (generation, pinning, and collapsing), an infinite canvas, and multi-user collaboration. Additionally, it allows cards to be generated dynamically through conversational syntax and supports embedded web applications.
Read Moredsh-sctrk: DeepSeek Harness (DSH) Session Trace Bar
dsh-sctrk is a visual session trajectory plugin for DeepSeek Harness. Using an 8px floating rail on the right, it maps long-session nodes to a color scale and automatically aggregates dense nodes into gradient cluster dots. Users can hover to preview nodes or click/drag the rail for precise jumps and proportional scrolling. The plugin is MIT licensed, supports theme adaptation, is easy to install, and is designed to address the difficulty of navigating long sessions.
Read Moredsh-greet: A Minimal Plugin for Registering the Greet Tool
The DSH Harness (DSH) plugin `dsh-greet` is a minimal reference implementation that demonstrates how to write a pure JavaScript plugin and publish it to GitHub. It requires no build step and registers the `greet` tool for direct agent invocation through `cordis.patch.yml` configuration. The plugin depends on `@deepseek-ai/dsh-tools` and `@deepseek-ai/cordis`, runs with the current process permissions, and is suitable for developers to quickly validate development workflows or integrate lightweight functionality. Installation and removal are both completed via the command line.
Read Moredsh-llm-failover: Model Provider Failover Plugin for DeepSeek Harness
`dsh-llm-failover` is a failover plugin for DeepSeek Harness (DSH), designed to resolve request interruptions caused by model provider rate limiting or quota exhaustion. It automatically detects `RATE_LIMIT` or `QUOTA` errors and switches to the next provider in the configured list until an available service is found or a permanent fallback is reached. The plugin supports provider-specific model configurations, cooldown mechanisms, and UI notifications, and can be configured via YAML files or settings cards. It depends on specific libraries, runs with the permissions of the current process, and is suitable for development scenarios that require high availability.
Read Moredsh-usage-guard: Usage Statistics Panel and Quota Guard for the DSH Web GUI
dsh-usage-guard is a usage monitoring plugin for DeepSeek Harness, designed to address cost control issues. It features silent background collection, session log backfilling, and real-time monitoring. Core capabilities include analyzing token usage by day and model, providing a dashboard with activity heatmaps, trend charts, and ring charts, supporting cost estimation with built-in DeepSeek pricing, and offering an optional daily limit guard with warn/block modes. It supports both Chinese and English, and can be installed via CLI and enabled from the settings menu.
Read Moredsh-notify: Universal Notification Plugin for DeepSeek Harness
The DeepSeek Harness (DSH) plugin `dsh-notify` addresses agent task status awareness by providing a layered notification solution. It supports in-app pop-ups and native system notifications, and features a Web UI settings panel and one-click operations for pending approval scenarios. Built on system-provided APIs for zero dependencies, it supports managing notification toggle settings and viewing logs. After installation, restart the DSH Web GUI. Note that native notifications serve only as reminders, clicking them provides no callbacks, and only Windows 10/11 and macOS are supported.
Read MoreDeepSeek Harness Plugin Advisor: Automatically Analyze Requirements and Recommend Plugin Installation
The DeepSeek Harness Plugin Advisor is a workflow plugin designed to solve the challenge of filtering agent plugins. It automatically analyzes requirements, checks locally installed plugins, searches the community ecosystem if needed, and scores candidate plugins using multiple signals (such as featured listings, stars, and download counts) to recommend the optimal option for one-click installation. The tool supports installation from GitHub or npm, includes a safety mechanism requiring user confirmation before installation, and its directory data has no official association with DeepSeek.
Read MoreDeepSeek Harness Plugin: dsh-llm-kimi
The `dsh-llm-kimi` plugin for DeepSeek Harness (DSH) provides a compatibility layer for Kimi models on the platform. It supports Kimi Code subscriptions and the Moonshot Open Platform API, covers domestic and overseas regions, and supports multimodal capabilities, tool calling, thinking mode, and hot updates. Users can configure API keys via the command line or a web interface. The plugin provides multi-routing support and strictly aligns with the official protocol for handling streaming responses and tool calls.
Read Moredsh-diffgeom: A Differential Geometry Symbolic Computation Plugin for DSH
The DeepSeek Harness plugin dsh-diffgeom integrates SymPy's differential geometry capabilities into DSH for handling geometry, physics, and tensor field problems. It provides support for manifolds, tensor fields, and differential operators (such as the exterior derivative d, the wedge product ∧, and the Lie derivative), along with Riemannian geometry computation tools (Christoffel symbols, Ricci/Riemann components). By invoking a Python backend, it enables direct manifold calculations in code and comes with 2D/3D Euclidean environments preset by default, allowing developers to quickly perform complex geometric computations without manually setting up their own scripts.
Read Moredsh-ai-teacher: Dedicated Teaching Agent Preset for DeepSeek Harness
The `dsh-ai-teacher` plugin provided by hatter123 adds an agent preset named “AI Teacher” to DeepSeek Harness. This preset objectively explains complex concepts in a step-by-step manner based on PDF documents, supporting Socratic guidance and an invisible motivation mechanism. Users only need to run the installation command to use the plugin in DSH for in-depth learning, without cumbersome prompt engineering.
Read Moreoh-my-deepseek-harness: Native Workflow Skill Set for DeepSeek Harness
`oh-my-deepseek-harness` is a modular skill set tailored for DeepSeek Harness (DSH), ported from `oh-my-codex`. It provides DSH with end-to-end tooling for planning, execution, quality assurance, and operations, supporting skills such as in-depth interviewing, consensus planning, strict autonomous delivery, and code review. The plugin is installed using the `dsh plugin` command and leverages DSH-native capabilities (such as `subagent`, `workflow`, and `ralph`) to enable structured agent collaboration. It requires file read/write permissions, is compatible with DSH 0.1.0-rc.6, and is available under the MIT license.
Read Moredsh-archived-sessions: Manage Archived Sessions in the Settings Panel
`dsh-archived-sessions` is an open-source plugin for DeepSeek Harness, designed to address the difficulty of locating archived sessions after they have been stored away. The plugin adds an “Archived sessions” section to the settings panel of a Web profile, listing hidden archived sessions. Users can restore a session by clicking its list item, ensuring that logs and workspace data are not lost. The plugin is released under the MIT license, requires restarting the profile to take effect, and currently supports Web profiles.
Read Moredsh-voice: A Turn-Based Voice Plugin for DeepSeek Harness Interview Scenarios
dsh-voice is a turn-based voice channel plugin for DeepSeek Harness (DSH), designed specifically for interview scenarios. It abstracts underlying engine details through Agent tools, supports independent selection of TTS/ASR engines, and provides speech synthesis, audio transcription, and health check functionality. The plugin supports cloud engines (Qwen, MiMo) and local engines (Sherpa-onnx, Whisper, etc.), with engines and parameters defined via a configuration file after installation. In an interviewer preset, the Agent can be instructed to use `voice_speak` to read questions and call `voice_listen` to wait for responses. The current version is v0.1, follows a turn-based-first strategy, and does not yet support full-duplex interruption or session persistence.
Read MoreDeepSeek Harness Plugin: dsh-uvm
`dsh-uvm` is a DSH client plugin maintained by Harzva, designed to manage Python virtual environments directly through the DSH web interface. It integrates full lifecycle management with `uv`, supporting environment creation and deletion, package installation, and command execution, and provides a `pip` fallback as well as read-only viewing of `conda` environments. The plugin offers capabilities such as project synchronization and dependency diff preview (M2), while strictly enforcing security restrictions: all operations are limited to the designated local directory, ensuring data privacy and system security. Users can install and use it with the official command.
Read MoreDeepSeek Harness Session Context Menu Plugin
The `dsh-session-context-menu` plugin provides Codex-style context operations for DeepSeek Harness session management without patching packages or querying private DOM. It supports pinning, marking as read, copying messages, and integrated actions such as renaming and archiving, invocable via right-click or keyboard shortcuts. State is stored only in a local JSON file and does not involve sensitive data, though it requires DSH to provide specific interface support.
Read MoreDeepSeek Harness Research Lab Plugin (dsh-research-lab)
`harzva/dsh-research-lab` is a standalone workflow plugin for DeepSeek Harness (DSH) that aims to address the difficulty of pinning down and reproducing experimental configurations. It provides a Manifest-first control plane that supports managing multiple inference runtimes (such as ReAct and Plan Guided) within the Web interface, and allows users to pin baseline parameters, generate ablation matrices, and output deterministic JSON configurations with checksums. The plugin does not invoke models directly; it is designed to help researchers rigorously compare and reproduce different reasoning strategies.
Read Moredsh-obsidian: Connecting Local Projects to Obsidian's Knowledge Graph
`dsh-obsidian` is a standalone plugin intended to address Obsidian's inability to directly scan code repositories. It uses DSH to scan local project metadata, generate corresponding notes and graph edges, and automatically surface knowledge associations between projects. Core features include native DSH integration, local project scanning, deterministic knowledge graph generation, dual transport modes, and incremental writes. After installation, users can configure environment variables, run scanning and graph generation in DSH, and view associations in Obsidian's graph view. The plugin runs entirely offline, is secure, does not affect manual notes, and does not require an additional REST plugin for basic functionality.
Read Moredsh-maclens: A DSH Toolkit for Running Apple's Vision Framework Locally
`dsh-maclens` is a local plugin for DeepSeek Harness that turns the native macOS Vision framework into a local tool. It requires no API key or network requests and supports core capabilities such as OCR, image classification, face detection, and document layout analysis. All pixel-level processing runs on the Mac Neural Engine, ensuring privacy and achieving sub-second speed. Installation is simple: on macOS 14 or later, it can be invoked via the CLI, providing agents with fully local visual-processing capabilities.
Read Moredsh-groupx: Bringing Local GroupX Multi-Agent Rooms to DeepSeek Harness
dsh-groupx is a standalone DSH plugin maintained by Harzva, designed to address tool fragmentation in local multi-agent collaboration. It bridges GroupX multi-agent rooms to the DeepSeek Harness (DSH) web interface, enabling status viewing, a shared timeline, and message routing. The plugin provides native tabs within DSH and, by default, connects to the GroupX Broker via a Loopback Host Bridge, supporting session views, message routing, and security validation. Installation requires starting GroupX first, then packaging and installing the plugin into the DSH Web Profile via CLI. The plugin supports Loopback connections only, ensuring data security and independence.
Read MoreDeepSeek Harness Plugin: dsh-control-plane
`dsh-control-plane` is a DSH plugin maintained by Harzva, designed to address the management challenges caused by multi-environment configuration confusion. It provides a local-first control plane that supports multiple runtime registration, session correlation, and web-based visualization. The plugin includes safety checks, pre-release validation, and extensible scenario adapters, with an additive integration design. Version v0.1.0 is currently released; developers should use isolated profiles to avoid installing local configurations into production environments.
Read Moredsh-cli-store: Secure CLI Tool Registration and Installation for DeepSeek Harness
dsh-cli-store is a DeepSeek Harness (DSH) plugin maintained by Harzva, designed to securely address installation and verification issues when agents invoke external CLI tools. As a registry, it provides a set of commands for searching, discovering, saving, installing, and diagnosing binary tools. The plugin follows a security-first design, managing installation flows with strict permission controls and validation mechanisms (such as dry-run mode and prevention of shell injection). Developers can use it to safely integrate tools like `gh` and `ffmpeg`, making it well suited to scenarios that require managing large numbers of external CLIs.
Read MoreDeepSeek Harness Plugin: dsh-agent-project-sync
The DeepSeek Harness plugin `dsh-agent-project-sync` is intended to consolidate project directories from agent tools such as Codex and Claude Code into DSH for unified management. It automatically discovers and registers these directories as native Workspaces, and provides CLI and DSH tools for scanning, inventory management, and secure synchronization. It is constrained by safety boundaries: it reads only file-header metadata, runs in dry-run mode by default, and never deletes existing data. After installation, users can manage the project inventory through the command line or built-in tools, while paying attention to the required permissions and the fact that the inventory contains local absolute paths.
Read MoreIntroduction to the DeepSeek Harness Image Understanding Plugin
`dsh-vision-analysis` is an image understanding plugin for DeepSeek Harness, designed to address the lack of visual capabilities in text-only models. It supports vision endpoints in OpenAI or Anthropic format, allowing integration with models such as Qwen or GPT-4o as well as local Ollama, and provides eight analysis modes, including OCR and chart data extraction. It is easy to install and accepts images as local files, URLs, or Base64 inputs. With image bridging configured, agents can paste images directly into conversations for analysis. The plugin prioritizes privacy by processing images internally only and returns structured JSON data for OCR and chart data extraction.
Read Moredsh-vision: A Plugin for Providing Vision Capabilities to DeepSeek Harness
dsh-vision is a vision plugin for DeepSeek Harness, designed to address the limitations of text-based models in image processing. By taking over the input stream, it supports image recognition and generation, with a multi-engine failover mechanism. Users can install it via the command line and use engines such as Codex and Gemini to read local images or generate images, with an interface that supports preview and download. The plugin follows strict temporary storage principles and offers flexible configuration options to manage credentials, engine parameters, and file retention policies, empowering agents with robust visual interaction capabilities.
Read More