dsh-web-recon: Web System Reconnaissance · Solidifying a Reusable Playbook
`dsh-web-recon` is a web system reconnaissance tool for DeepSeek Harness. To address the high cost and poor stability of traditional screenshot-based automation, this plugin uses the CDP protocol to capture real HTTP interfaces and accessibility trees, distilling them into reusable “playbooks.” Its core features include dependency-free operation, interface summarization to save tokens, and isolated browser instance management. Users run `web_recon` to scout a page, then use `web_replay` to directly reuse the captured interfaces, eliminating the need for repeated screenshots or clicks. The plugin supports manual login authentication and stores data as JSON files, greatly improving the efficiency of web automation development.
Read Moredsh-usage-info: DSH Local Usage Statistics Plugin
dsh-usage-info is a localized usage statistics plugin for DeepSeek Harness (DSH). It adds a "Usage Info" item to the Web interface, enabling users to view today's and this month's API call count, token consumption, and cache hit rate. Core features include local persistent storage, automatic incremental synchronization, grouping and sorting by vendor/model, and privacy-safe data processing. Installation is straightforward and requires no database, as it relies entirely on local session log analysis, making it ideal for developers to monitor costs and invocation logic.
Read Moredsh-subagent-model: Independent Model and Preset Configuration for DSH Subagents
The DeepSeek Harness plugin is designed to address the resource overhead caused by a main agent invoking subagents that inherit the main model. It adds a "Subagent Model" menu, allowing users to uniformly configure independent providers, models, and reasoning effort for all delegated subagents. The plugin provides official UI integration, a globally applied delegation policy, and preset tier management, while supporting foreground waiting and context inheritance. Installation requires mounting an npm package and configuring dependencies; configuration is stored in a local JSON file and is designed for the DSH Web environment.
Read MoreDeepSeek Harness Plugin: Replacing the Hero Title and Applying an Artistic Text Style
This plugin aims to provide highly customizable Hero section styling for the DeepSeek Harness empty-state page. It takes over the default title “Explore Uncharted Terrain” through DOM manipulation, allowing users to replace the copy, hide the “Preview” label, and apply artistic text effects (such as gradients, outlines, and shadows). It can also fill the DeepSeek little whale icon with the official blue and features a Monet color-picker integration that extracts the dominant color from the background to generate a gradient. After installation, users can easily adjust the copy and style configuration in `client.js`.
Read MoreWooden Fish Pet Plugin Overview
Wooden Fish Pet in the DeepSeek Harness plugin system is a desktop pet that provides visual feedback by simulating strikes and is used to record daily counts. Maintained by ShenpingDD, it supports dragging, displaying today’s count on hover, and triggering an animation on click. Data is aggregated by date and can be viewed as a heatmap. Its position and state are persisted, and it automatically adapts to DSH themes. Installation requires running a specified command in the terminal and takes effect after restart. The plugin runs entirely locally without network requests and stores data in local files; however, the current version offers visual feedback only and does not support audio feedback.
Read Moredsh-evidence-arena: A Multi-Model Programming Comparison Workbench for DeepSeek Harness
`dsh-evidence-arena` is a community plugin for DeepSeek Harness that addresses the lack of evidence and cumbersome switching in multi-model parallel programming comparisons. It supports running two or three Builders in parallel within an isolated environment to solve the same task, uses an independent Reviewer to perform logic and security checks, and provides reproducible checkpoints and per-file diffs. The plugin emphasizes an evidence-first approach, writing back only human-verified code to the original workspace, and operates in a zero-tool sealed environment to ensure safety. It is compatible with DSH version 2026-08-19 and is suitable for model performance evaluation and code generation quality validation.
Read Moredsh-rss-for-shadow-suite: An All-in-One RSS News Subscription Plugin for the Shadow Suite
dsh-rss-for-shadow-suite is a subscription plugin for DeepSeek Harness (DSH), designed to address the cost for developers of managing multiple RSS sources. It integrates 33 preconfigured sources and includes two scheduled routines: an AI daily briefing and an arXiv papers digest, supporting automated scheduling and centralized distribution. Users need to install web and headless profiles separately and manually configure the output directory (such as an Obsidian path). Proxy-based access is supported, and files are stored locally. This plugin lowers the configuration barrier for information aggregation and is suitable for developers who need to automatically obtain cutting-edge technology, AI research, and computer science papers.
Read Moredsh-course-logic-extractor: Course Logic Extractor
`dsh-course-logic-extractor` is a DeepSeek Harness plugin designed to algorithmically reconstruct the “what to teach first, what to teach later” design logic of a course or paper. It offers three input modes (direct analysis, paper mode, and autonomous material sourcing), supports intent-based entry routing, and can process non-text materials such as PDFs through a pipeline. The plugin can generate a logic chain analysis report, a learning roadmap, a process canvas, along with an evidence layer and a material layer. It requires configuring a Zhipu OCR key and features local modification protection and portability. During execution, it generates a complete local directory structure containing planning, reports, evidence, and materials.
Read Morellm-deepseek-image-admit: DeepSeek Image Admission
DeepSeek's official model metadata restricts image admission. This plugin takes over the official routing and modifies the metadata—completing `inputModalities` to `['text', 'image']`—to bypass the harness's image blocking and allow image messages to pass admission pre-checks. It relies on the official artifact layout, transmits no images via the API, and must be used with plugins such as `image-bridge` to transcribe image content during the `pre-step` stage, without changing user behavior. The installation command is `dsh plugin --profile web add github:Seryta/llm-deepseek-image-admit`.
Read Moredsh-plugin-safety: Plugin Health Alerts and Automatic Repair
The DSH plugin `dsh-plugin-safety` is designed to address failure management issues in the plugin ecosystem, providing the final two links in fault prevention and control: pre-startup checks and runtime automatic repair. With zero third-party dependencies, its core features include automatically disabling the plugin and persisting alerts to disk when prechecks fail, persistently displaying alerts in the browser, and automatically creating a headless repair session on the Host side to attempt recovery when alerts are detected. The system uses file locks to prevent concurrent duplicate startup, with alert data stored in `notice.json`. After installation, `dsh-web` must be restarted. The precheck script can also be integrated into systemd's `ExecStartPre` to perform blocking or non-blocking checks before startup, enabling automated health maintenance for the plugin.
Read MoreDeepSeek Harness Plugin: dsh-writing-shape Writing Shaping
`satan9394/dsh-writing-shape` is a DeepSeek Harness (DSH) plugin designed to establish writing discipline through “writing shaping” and “turning raw materials into text.” It belongs to the model reasoning category and supports converting source materials into structured text. Users can install it from the DSH directory page or the GitHub repository. When using it, users should check the source code for security.
Read MoreDSH Plugin: Track Driver Development
The Track-driven development feature of the DeepSeek Harness (DSH) plugin aims to improve agent development efficiency. By integrating specifications, execution plans, and quality gates into the development process, it enhances the consistency and controllability of the workflow. The plugin supports a Track-driven development mode, inspired by Agents Conductor. Since installation and detailed usage rely on the GitHub repository documentation, users are advised to consult it directly. When using the plugin, review the source code for security and verify license compliance.
Read Moresatan9394/dsh-rag: RAG (Retrieval-Augmented Generation) Skills under DeepSeek Harness
`dsh-rag` is a RAG skill plugin for DeepSeek Harness (DSH), designed to improve an agent's information accuracy and response quality through retrieval-augmented generation. The plugin provides core capabilities such as vector stores and embeddings, document pipelines, retrieval strategies, and hallucination control. Since the installation command is missing and no examples are provided, users need to check the GitHub repository themselves for functional details.
Read MoreIntroduction to the DSH Plugin: satan9394/dsh-quant-backtest
`satan9394/dsh-quant-backtest` is a quant backtesting skill for DeepSeek Harness (DSH), designed to improve strategy development efficiency. The plugin integrates quant backtesting, risk metric calculation, and bias avoidance features, supporting walk-forward and Sharpe drawdown analysis. Developers can use it directly in agent workflows, but should pay attention to permissions and licensing. No official installation command is currently provided; it is recommended to review the source code.
Read Moredsh-python-development: Python Development Patterns and Structural Type Safety
DeepSeek Harness (DSH) has released the `dsh-python-development` plugin, designed to provide a structured development environment and type-safe support for Python. Drawing on experience from mature projects, it helps developers address issues of loose code and difficulty in ensuring type safety. Users should consult the GitHub repository for installation and usage instructions. Because license information has not been disclosed, be sure to check the source code license before use.
Read Moredsh-prompt-engineering: DeepSeek Harness Prompt Engineering Patterns
The `dsh-prompt-engineering` plugin, developed by `satan9394`, is an extension for DeepSeek Harness designed to streamline prompt creation. It provides standardized prompt engineering patterns, supports CoT, few-shot examples, and template reuse, and helps improve debugging efficiency. Since no official installation command is provided, users should refer to the community directory or GitHub for guidance. During installation, proper permissions and security compliance must be considered.
Read MoreDeepSeek Harness Plugin: satan9394/dsh-prompt-audit
`satan9394/dsh-prompt-audit` is a plugin for DeepSeek Harness that aims to address prompt auditing, model migration, and outdated pattern scanning. The plugin runs with the permissions of the current process, making it suitable for developers who need fine-grained control over prompt logic. Currently, no official installation command is provided; refer to the source code and license.
Read Moredsh-pptx-engineering: PPTX Generation and Validation Skill
`dsh-pptx-engineering` is a plugin in the DSH ecosystem for engineering-oriented processing of PPTX documents, using the `pptxgenjs` library to implement generation and validation. The plugin is licensed under Apache-2.0 and is inspired by Anthropic skills. Currently, no official installation command is provided; due to missing documentation, developers need to check the source code themselves to ensure environment compatibility.
Read MoreDSH Plugin: satan9394/dsh-ppt-creator (Chinese PPT Generation)
`satan9394/dsh-ppt-creator` is a DSH plugin designed to address the cumbersome layout challenges in generating Chinese presentation slides. It converts text outlines into visual designs, supports exporting to HTML or PPTX formats, and provides a layout self-check checklist. The plugin runs with the permissions of the current process, so its source code and license compliance should be reviewed before use. Since no official installation command is provided, it is recommended to consult the specific documentation and installation instructions in its GitHub repository or plugin directory.
Read Moredsh-postmortem: Incident Postmortem Writing Assistant
The DSH plugin dsh-postmortem is designed to address the time-consuming and off-topic challenges of writing incident postmortem reports, with a focus on generating standardized documentation. Through structured guidance, it helps users organize timelines, identify root causes, and formulate action items, while emphasizing a blameless postmortem culture. The plugin runs with the permissions of the current process, so security care is needed when sensitive data is involved. It is suitable for technical teams that require standardized processes and want to build a repository of lessons learned; refer to its GitHub repository for installation instructions.
Read Moredsh-ponytail-dev: Lazy Senior Development Philosophy and Minimal Discipline
The DSH (DeepSeek Harness) plugin dsh-ponytail-dev aims to provide a “lazy senior development philosophy,” combining a YAGNI ladder strategy, root-cause fix logic, and minimalist discipline. The core feature list is currently empty, and because the README fetch failed, the exact installation command and typical usage remain unconfirmed. Before use, review the source code and license, and run it with the permissions of the current DSH process. The plugin focuses on development discipline; the related source code and directory page can be accessed via the provided links.
Read Moredsh-planning-files: Three-File Persistent Planning Plugin
The DeepSeek Harness (DSH) plugin uses a three-file persistence mechanism to prevent task plans, discoveries, and progress from being lost during process restarts or session switches. Maintained by satan9394, the plugin supports managing long-lived planning sessions and must run with the permissions of the current process. It is recommended to review the source code and license before installation. For more details, refer to the GitHub repository or plugin directory.
Read Moredsh-planning: DeepSeek Harness Task Planning and Decomposition Plugin
dsh-planning is a DeepSeek Harness skill plugin designed to address the challenges agents face when handling complex tasks, such as vague objectives and the difficulty of decomposing and validating tasks. The plugin provides structured task management capabilities, supporting the breakdown of vague goals into verifiable subtasks, dependency ordering, difficulty estimation, and milestone setting. It runs with the permissions of the current DSH process and is suitable for complex workflow development scenarios. For specific installation and usage instructions, refer to the source code in the GitHub repository.
Read Moredsh-personal-content-discovery: Cross-Platform Personalized Content Discovery
`dsh-personal-content-discovery` is a DSH skill plugin maintained by satan9394, designed to address the challenge AI agents face in obtaining cross-platform personalized content. By constructing a five-tier soul profile, inferring MBTI personality types, and integrating cross-platform interest data, the plugin enables local-first, precise content discovery and proactive exploration. It supports recommendations from a simulated friend's perspective and is suitable for agent developers who need to retrieve and analyze personalized content across multiple platforms.
Read MoreDSH Plugin: satan9394/dsh-persistent-memory — Cross-Session Persistent Memory Compression
satan9394/dsh-persistent-memory is a memory plugin for DeepSeek Harness, designed to address the challenges of difficult cross-session memory retention and high token consumption for agents. Inspired by the design of claude-mem, it provides cross-session compressed storage, lifecycle hook capture, three-tier retrieval (search/timeline/details), and progressive disclosure. The plugin supports private tags and hybrid search to significantly reduce token usage. There is currently no official installation command, so users must manually access the plugin directory or GitHub repository to deploy it.
Read More