All-in-One Operations Console for Solopreneurs: Ready-to-Use Coverage for Topic Selection, Social Posting, Live Streaming, Pricing, and Weekly Reports
The "One-Person Company Ready-to-Use Operations Console" (@user_ff29c4df/opc-all-in-one) is a community-sourced Agent Skill, version 1.2.0, categorized under business-ops, published by user_ff29c4df, and requires no additional API key. It is designed for the full spectrum of one-person company operations, covering five task types—topic selection, social media post writing, livestreaming, pricing, and weekly reports (excluding tax filing and bookkeeping)—and produces ready-to-use deliverables such as a topic library, social media copy, livestream scripts, pricing and product lineup, and weekly report templates. The official description emphasizes compliance with platform rules, accurate timeliness labeling, and non-fabrication of data. To use it, simply tell it what you sell, and you will receive directly applicable content packages. The article also outlines the target audience, usage limitations, and the approach of referring to the official page for installation instructions, as public materials do not provide installation commands.
Read MoreBusiness Analysis & Decision Making | Dun Shi Mu: A Combat Simulation for Major Business Decisions
This article introduces the community Skill '商业分析决策｜盾十目' (Business Analysis Decision | Shield Ten Eyes) available on SkillHub (version 1.1.1, category: business-ops, no API Key required). It treats major business decisions as war-game simulations, executing heuristic questioning, resource guidance, three-stage intelligence gathering, and multi-role adversarial exercises following the STORM-Pro v1.1 process, followed by structured analysis using SWOT, sensitivity, and stakeholder matrices, then butterfly effect precedent verification and red team audits, and finally outputs CEO briefings and circuit-breaker recommendations, with the ability to generate interactive HTML reports or McKinsey-style Word documents. The article explains in order its positioning, nine core capabilities, available installation and trigger methods, reproducible typical use cases, as well as applicable scenarios and limitations—including cautions such as its community origin, zero installation count, simulation quality depending on input materials, and possible evidence gaps in precedent verification—and provides the official address.
Read MoreSeewo Whiteboard Teaching Icon Extension Library: Chinese Retrieval and Export of 818 SVG Icons
This article introduces a Seewo Whiteboard teaching icon extension library. Released by Jishang Network Technology, the Skill targets Seewo Whiteboard 5 courseware creation and provides 818 icons, 21 categories of Chinese-language search, fully embedded vector SVG assets, and one-click export of transparent-background PNGs from the web page. Its official page is on SkillHub at @jishangtech/js-96, version 1.0.0, source: enterprise, category: education, tags: pricing_type=paid and requires_api_key=false, meaning it is a paid skill but does not require an API key. The article lists verified metadata and notes that installation commands and reproducible examples were not provided in the official materials and were not fabricated. It further cautions that the pricing model (a one-time purchase of ¥4.99 with unlimited use, plus one free use per IP) and the boundaries regarding copyright and commercial use should be confirmed with the publisher, and that ratings and download counts should be taken as per the official page.
Read Moredsh-weknora-archive：将 DSH 会话增量归档到自托管 WeKnora
dsh-weknora-archive is a DSH host-side plugin that incrementally archives local sessions to a self-hosted WeKnora instance, serving as a writing complement to the official read-only dsh-weknora. Version v0.4.0 polls parsing status post-upload, writes to archive-manifest.json, and can generate a local Three.js star map for ingested data. It supports tools such as auto, session, and status, as well as environment variable configuration and scheduled full archives. Licensed under MIT, it requires Node.js >= 24 and the DSH web profile.
Read More50 Years of Classical Study of Orthodox Tradition · East-West Folk Culture Reference Consultant: A Folk Culture Skill Dedicated to Literary and Historical Popularization
This article introduces a Skill for folk culture consultation, strictly limited to the scope of popular science regarding the origins of literature and history. It simulates a senior scholar with over fifty years of experience in studying orthodox classical texts. Based on the theoretical foundations of the Five Arts of Taoism (Mountain, Medicine, Life, Divination, and Physiognomy), it covers Bazi, Zi Wei Dou Shu, Qi Men Dun Jia, the Chinese Zodiac, Xuan Kong auspicious directions,auspicious date selection, Yang宅 folklore layouts, as well as interpretations of the twelve zodiac signs and Tarot symbolic psychology. Its core boundary is strict: it does not cast horoscopes, tell fortunes, analyze fate charts, or predict fortunes or misfortunes; it focuses solely on the popular science of Taoist philosophy and traditional folk culture. Version 1.0.2 of the Skill, categorized under knowledge-management and sourced from the community, was published by user_f662ce5f. It requires no additional API Key and includes built-in quick-start guides, best practices, FAQs, and complete Q&A examples; it actively asks follow-up questions if information is incomplete. The article provides installation and activation commands, typical usage scenarios, and limitation descriptions, making it suitable for developers seeking a stable and restrained cultural explanation assistant.
Read MorePPT Design: A Skill for Generating and Beautifying Professional Presentations
This article introduces a technical overview of the Community Edition PPT Design Skill (@user_db7424f5/udb7-ppt-design, version 1.0.0, categorized under office-efficiency). Beginning with common pain points encountered when integrating PPT capabilities, the article clarifies the Skill's positioning and release information. It outlines three verified core capabilities: multi-industry adaptation, support for diverse styles and visual systems, and trigger word entry points for design, beautify, improve, and create. It also highlights the `requires_api_key=false` feature, which ensures a low barrier to entry for integration. The installation section realistically notes that specific platform commands were not provided in the available documentation, offering only the official address and the SkillHub page instead. Typical usage section organizes three types of invocation examples based on trigger words, emphasizing the three essential elements of industry, style, and page count. Finally, the section on applicable scenarios and limitations lists important considerations such as gaps in documentation, early-stage versions, and zero installation volume.
Read Moredsh-request-header-fix: Add x-opencode-session to custom model requests
DSH only automatically adds x-opencode-session for some providers out of the box. If your self-hosted or third-party endpoints require this header, you can use the community plugin dsh-request-header-fix. In the web profile’s settings, enable it per provider and model; it injects the current session UUID only when llm/stream is hit. It is disabled by default and only modifies the profile. Based on the repository’s README, this article covers the installation command, the configuration field headerFix, how to restart dsh web, and usage notes.
Read MoreWeChat Channels Download (including free download channels): Parse WeChat Channels shared links into local MP4 files.
Introducing the WeChat Channels (Video Account) Download Agent Skill (including free download channels): Simply paste a WeChat Channels share short link to parse the title, author, cover, and topic tags, and download the original quality, watermark-free MP4 video to your local device. It supports searching for creators by nickname, listing work short links, batch converting to direct links for download, and audio extraction. The tool runs locally with zero configuration and zero dependencies, requiring no packet capture tools or Playwright.
Read MoreNovel Writing, Dream Shadow Clone MYavatar: Full-Process Novel Creation Skill from Idea to Text
"Novel Writing, Mengying Avatar" is a Skill maintained by the publisher Mengying Fenun (username: Mengyingavatar), located at @mengyingavatar/myavatar, version 1.0.3, and categorized under content creation. It breaks novel creation into a complete pipeline: brainstorming -> characters -> worldbuilding -> outline -> volume outline -> chapter outline -> manuscript. The pipeline can be customized by male/female-oriented channels, thematic direction, core elements, and writing style, with style settings persisted. It automatically reads and writes local materials and recursively backfills missing files to maintain consistency across long-form works. In addition to long-form novels, it supports medium and short stories, novel adaptations, AI book breakdowns, novel-to-script conversion, short dramas, medium/long-form dramas, and script doctoring. Trigger phrases include "Mengying Avatar," "write a novel," "write a short drama," and "break down a book." requires_api_key=false, but it is a paid Skill. This article was written based on gathered materials; the official page was inaccessible, and no install command was provided. Rating: 100000; Downloads: 1127.
Read Moredsh-turn-notify: Windows System Notification at the End of the Current DSH Session Turn
The DSH client plugin `dsh-turn-notify`, maintained by 99galaxy, sends system notifications on Windows when the current session turn ends, indicating the conversation name and optional elapsed time. Clicking the notification attempts to bring the DSH window back to the foreground. It supports foreground Do Not Disturb mode, skipping sub-agents, and title sanitization. Install the npm package or GitHub source via `dsh plugin --profile desktop`, and configure settings such as `enabled` and `notifyWhenFocused` in `cordis.patch.yml`. DSH must remain running, and system notifications must be permitted. The "Turn Completed" notification does not differentiate between success or failure. MIT 1.0.4.
Read MoreE-commerce Main Image & Detail Page Diagnosis Pro: An Evidence-Based Health Check for Product Listings
This community skill, published by user_e52351b3 (version 1.3.0, categorized under business-ops), does not require an API Key. It focuses strictly on diagnostics rather than content generation: by uploading screenshots of existing main images, detail pages, search results pages, or product pages into the conversation, it outputs per-image observations, confidence-based heuristic scores, core issues, and prioritized P0/P1/P2 improvement checklists. It covers main image diagnostics, detail page analysis, click potential analysis, competitor comparison, and mobile device checks. The article clarifies its boundaries compared to copywriting-focused skills: when users need to rewrite copy, generate storyboards, or create execution drafts, they should switch to the ec-visual-copywriting-pro skill. It also reminds users not to treat these heuristic scores as actual CTR/CVR metrics or real sales traffic data.
Read Moredsh-seed: Personality and Memory Grown from Experiences
dsh-seed is a memory plugin for the DSH community that preserves designations, boundaries, and taboos during turn-end extraction, governance, reflection, and self-description reviews, while updating its self-identity based on evidence. Based on the official README, this article details the installation commands, cordis configuration, data directories, and panel ports, while highlighting注意事项 regarding the plugin's execution under DSH process permissions, self-learning mechanisms, and LLM configuration.
Read MoreMember Operations - A Practical Perspective: Clarifying the Data Foundation for Operational Actions
Membership operations teams frequently encounter issues such as inconsistent metric definitions and a lack of data-driven foundations for operational actions. This Skill (Membership Data Consultant · Majia Practical Edition, majia-huiyuan) covers membership metric definitions, RFM analysis, formulas for repurchase/retention/churn, redemption rates, average order values, member segmentation/grouping, audience targeting, tagging systems, CDP, OneID, Cohort analysis, membership data warehouses (DIM/DWD/DWS/ADS), SQL/DDL, data dictionaries, data quality, membership dashboards, and replication of Guanyuan BI. When users propose operational actions such as user recall, frequency increase, churn prevention, churn warning, new customer conversion, channel migration, or guide task assignment, this Skill provides the data rationale behind these actions (target audience, timing, intensity, assignee, and tracking methods). For execution content (Moments, mass messaging, welcome语, community SOPs, WeCom operations) and comprehensive private domain diagnostics, please refer to majia-siyu. All numerical values are simulated; only the structure and metric definitions are for reference.
Read MoreKunlun Growth · WeChat Official Account Article Writing: Introducing Structured Output to Article Writing
This article introduces the Agent Skill "Kunlun Growth · WeChat Article Writing" (@kunlungrowth/kunlun-wechat-article-writer, v1.2.0, enterprise source, categorized under content-creation). This Skill solidifies the repetitive structure of WeChat article writing into a fixed output format consisting of 15 title candidates, a body skeleton, 3 types of openings and 3 types of closings, layout parameters, and image placeholders. Trigger words include WeChat articles, tweet writing, soft ads, graphic layouts, and viral articles. No API Key is required. The text explains its positioning, verified core functions, official address, and enablement methods, while providing applicable scenarios and precautions: it outputs structure and candidates rather than finished drafts; given its installation count is 0 and community verification may be limited, it is recommended to run a trial first.
Read MoreData Analysis Skill Pack DataMaster Pro: Integrates Scraping, Cleaning, Visualization, and Reporting into One Pipeline
DataMaster Pro is a data analysis Skill developed by RealQ, released as an enterprise resource at coordinates @org-eyw2ohcx/gdp6539-data-analysis-skill-100. Classified under data analysis with version 1.0.0, it covers a four-step workflow: automated data extraction, cleaning, visualization, and report generation. Targeted at data analysts and operations personnel, it aims to replace manual Excel processes. Available data indicates a rating of 100,000 and 1,076 downloads, with zero installations, tagged as `requires_api_key=false`. However, installation commands, specific usage instructions, and toolchain details are absent from the documentation, and verification via the official page is unavailable. Therefore, this article solely states verifiable positioning and capability boundaries, recommending a small-scale validation and confirmation of internal compliance requirements prior to adoption.
Read MoreIndustrial Internet Expert: Entrust Factory Digital Planning and System Selection to a Skill
The "Industrial Internet Expert" is a community-sourced skill in SKILL.md format, published by user_83002a29. It is categorized under "professional" (Version 1.0.0) and requires no API Key. It activates in response to four types of requirements: factory digital transformation planning; ERP/MES/APS/PLM/WMS/QMS/SCADA/IIoT/digital twin/AI system assessment; manufacturing pain point diagnosis; and IT/OT integration guidance. It offers seven core capabilities: current state diagnosis and requirement analysis, full-stack industrial system planning, cost-effectiveness solutions (self-built/SaaS/customized), phased implementation planning, supplier selection and evaluation, implementation risk management, and AI/ML industrial application deployment. It covers discrete manufacturing, process industry, automotive, electronics, food and beverage, and chemicals. This article provides a verified positioning, the seven core capabilities, reproducible examples organized by official activation conditions, and usage notes regarding unverified installation methods and inconsistent metric sources.
Read Moredsh-cloudflared-manager: Manage Cloudflare Tunnels and Local Services in DSH
dsh-cloudflared-manager is a DSH workflow plugin published by 96bearli that manages Named tunnel ingress and DNS within the panel, as well as local service start/stop operations, log viewing, source code editing, and trycloudflare temporary public exposure. It supports installation via npm or GitHub, uses $DSH_HOME/cloudflared-setup/ as the default working directory, and provides POST /dsh-cf/api. The current version is designed for Linux x64 container environments; review the source code before installation, and do not commit credentials files to git.
Read Morevscode-pet: Integrate the official VS Code Blobby chat pet into the DSH web input box
vscode-pet (96274zxc, v1.2.27) ports the VS Code official Blobby's sprite sheets, frame timing, and interaction rules to appear above the DSH web input box, supporting chat reactions, drag-and-throw, 13 official hats with color customization, and a jumping mode; on Windows desktop, optional desktop display next to the taskbar is available. The installation command is `dsh plugin --profile web add github:96274zxc/vscode-pet`. This ported version does not track achievements, and settings are stored locally in localStorage. MIT licensed; not an official Microsoft product.
Read MoreProject Templates: Collection of New Project Scaffolding Templates
The project template (@org-02qudk26/templates, v1.0.0) is a collection of new project scaffolding templates released by Zhiyu Technology, categorized under dev-programming and based on the universal SKILL.md format. The article first explains that it addresses the problem of having to repeatedly set up directories, configurations, and conventions whenever a new project is initiated. It then clarifies that this is merely a template collection rather than a runtime tool, and the core functionality section only documents verified positioning information. The official documentation does not provide installation commands, directory paths, configuration items, or reproducible examples. The article does not fabricate these details but instead provides steps for verification via the official address. It also notes that the differences in how Cursor, Codex CLI, and Claude Code load the templates cannot be verified, so no separate assumptions are made. Important notes include: the templates do not contain business logic; the anomalous rating magnitude of approximately 100000 requires independent judgment; there are inconsistencies in the sources of download and installation data; and the long-term maintenance status of the enterprise source is unknown. The article concludes by providing the official address for users to verify before use.
Read Moredsh-session-dispatch: Dispatching Subtasks into Independent Workspace Sessions
The DSH plugin dsh-session-dispatch (0.1.0, MIT) exposes Session Controller capabilities as five model tools: conversation_create, send, wait, read, and list. The model can create new top-level workspace conversations, send commands, wait for turns to complete, and read replies. Subtasks run in independent persistent sessions rather than as subagents. It applies to the Web profile; the installation command is `dsh plugin --profile web add dsh-session-dispatch`, and DSH must be restarted after installation. Compatible with DSH 0.2.x and Node.js 22+.
Read MoreAccessing Free NVIDIA Models: Integrating Large Models from build.nvidia.com into WorkBuddy
A community-sourced Skill (Coordinates: @user_f9b838d6/nvidia-free-models-workbuddy, Version: 1.0.0, Category: ai-agent, Tag: requires_api_key=true) integrates 100+ free large models from NVIDIA build.nvidia.com (including Kimi K3, DeepSeek, MiniMax, Step, Nemotron, Mistral, etc.) into WorkBuddy as custom models. The free tier does not consume WorkBuddy credits, with a rate limit of 40 requests/minute. It covers the entire workflow: registering or reusing API keys, fetching the official real-time model list, writing to models.json, and performing individual verification tests. It also provides troubleshooting solutions for three specific pitfalls: models going offline (404/410 errors), proxy environment anomalies, and null content returned by reasoning models. Trigger keywords include: NVIDIA free models, build.nvidia.com, nvapi, free LLM API, integrate free models, save credits, and model offline errors.
Read MoreData Table Expert: A Table Processing Assistant for the General Public
Data Spreadsheet Expert is a community-sourced office-efficiency Skill located at @user_92725383/data-spreadsheet-expert. Version 1.0.0, published by user_92725383, it does not require an API Key for activation. It is designed for high-frequency, rule-based general spreadsheet organization needs. Verified capabilities include Excel/CSV format conversion, data filtering, sorting, deduplication, merging, multi-sheet association and matching, basic statistics (sum, average, count), and batch spreadsheet processing. The official description lists trigger scenarios such as "processing spreadsheets," "merging Excel," "filtering data," "deduplication," "data statistics," "CSV to Excel," "multi-sheet merging," and "spreadsheet organization." The article covers what it is, core functions, installation and activation methods (note that the official page body could not be scraped, so unverified installation commands and configuration items are omitted), applicable scenarios, and necessary limitations. Limitations include the exclusion of pivot tables and complex modeling, the requirement for clear association keys in multi-sheet relationships, the need to specify statistical scopes in requests, and the recommendation to check the source page for maintenance status before use. The official URL is https://api.skillhub.cn/user_92725383/data-spreadsheet-expert.
Read MoreLearning Analytics Assistant: Transforming Scattered Learning Signals into a Diagnostic Framework
The Learning Analytics Assistant is a community-sourced, general-purpose Skill (Coordinate: @user_af163eaa/z-learning-analytics, Publisher: user_af163eaa, Category: Education, Version: 1.0.1) designed to address a key pain point for education-focused developers: possessing structured data such as grades, error logs, and study habits, yet struggling to derive actionable diagnostic conclusions. Instead of merely providing total scores, it aggregates multiple learning signals to pinpoint knowledge weaknesses and learning styles, offering tiered tutoring recommendations. Activated by trigger phrases like "Analyze learning status," "Where is this student weak?" or "How to improve grades," the tool does not require external API keys. This article outlines its core functions, installation and activation methods, typical input/output usage, applicable scenarios, and caveats (notably that it does not handle data collection or statistical charts, input quality determines diagnostic accuracy, and conclusions require human review), along with the official reference address.
Read Moredsh-preset-bridge: Connecting SillyTavern Presets and Character Cards into DSH Normal Sessions
dsh-preset-bridge is a DeepSeek Harness client plugin that integrates SillyTavern's chat-completion preset JSON and PNG/JSON character cards into regular chat sessions. It supports per-block toggling and editing, My Settings, Basic Character, a master switch, and independent per-session configuration, all managed via a floating panel and an HTTP API. Install it as a local link with `dsh plugin --profile desktop add <absolute repository path>`. MIT-licensed; the author notes it has only been tested on standalone Windows, and the source code should be reviewed before use.
Read MoreMemory Radar: Basic Security Scanning for Agent Memory Files
Memory Radar (memory-radar-free) is a fundamental security scanning tool designed for AI Agent memory files. Coordinates: @u_8ee34d77/memory-radar-free; Version: 1.0.1; Category: it-ops-security. It is maintained by community publisher u_8ee34d77, and the page indicates `requires_api_key=false`. It offers three core capabilities: detection across seven major threat categories, five-level security classification, and local mode scanning, covering common threats such as malicious instructions, Prompt Injection, and credential leakage. Scans run locally with zero network requests; the foundational LLM is provided by the Agent platform rather than bundled within the tool itself. Typical use cases include periodic memory security audits, injection detection after importing external data, and credential leak troubleshooting. The core philosophy is "scan before use." The official documentation does not provide command-line installation steps, nor does it offer specific configuration instructions for Cursor, Codex CLI, or Claude Code. This Skill is based on the universal SKILL.md format, and activation depends on the SkillHub page and the Agent platform's loading mechanism. It is important to note that this tool provides only basic detection with limited coverage. Classification results require manual review and handling. It is suitable for long-running Agent projects that involve the ingestion of external data.
Read More