WeChat Official Account Article Publishing: A Quick Look at a Community Skill
This introduces the community Skill "WeChat Official Account Article Publishing" on SkillHub. Located at @user_5c92a6a1/xiaoshitou-wechat-publisher, it is version 1.0.0, categorized under content-creation, and tagged with requires_api_key=true. The retrieved data does not provide installation commands or usage examples, so no commands will be constructed. The text points out an inconsistency between the name and the description: while the name refers to publishing articles for Official Accounts, the description mentions "Volcengine/Other AI Image Generation API Key." Therefore, actual capabilities must be verified by examining the source code. At the time of retrieval, the page data showed a rating of 100,000, 893 downloads, and 0 installations; please refer to the real-time page for the most current statistics.
Read MoreDSH X: Delete Session in DeepSeek Harness
DSH X (dsh-x) is a DeepSeek Harness plugin maintained by Linzzz14, targeting the web profile. It currently supports deleting individual sessions and sub-sessions from the session menu, cleaning up disk records, workspace ownership, and caches, without removing shared attachments. The article provides official installation commands for the client and CLI, outlines deletion scopes and limitations (deletions are irreversible and impossible while running), and emphasizes the need to review the source code and the MIT license before installation.
Read Moredsh-mouse-blur: Background blurs on mouse movement and clears on rest.
dsh-mouse-blur is a DSH Web client plugin maintained by linhaiyutuo (npm package dsh-mouse-blur; version 0.3.0; MIT license). When background images or videos are configured via dsh-cool-theme, the background blur returns to baseline automatically after the pointer has been still for approximately 1.2 seconds; moving, clicking, or scrolling restores the blur at a configurable percentage rate (default: 40% of the full-scale 24 px). Settings are located under “Theme · Mouse blur”, allowing configuration of active blur, the transition time, and the still-duration time. Installation command: dsh plugin --profile desktop add dsh-mouse-blur. Compatible with cool-theme 0.9/0.10; given-elt verified on DSH 0.2.1-rc.2 with cool-theme 0.10.0.
Read More@lingxinsuixing/dsh-agnes: Invoking Agnes in DSH for Image-Text-to-Video Generation
Introducing the community plugin @lingxinsuixing/dsh-agnes: It registers Agnes text-to-image, text-to-video, and optional image recognition tools within DeepSeek Harness. Tool results return as plain text (containing paths and URLs), and media files are written to a local directory. This covers the official installation command, the configuration order for AGNES_API_KEY, parameters for the three tool types, default and free models, output directory and saveTo constraints, as well as integration considerations such as avoiding read_image usage to prevent HTTP 400 session errors.
Read MoreSix Sigma DMAIC in Practice: Turning the Five Stages into a Conversational Operational Coach
Six Sigma DMAIC in Practice is a community Skill that provides structured, hands-on guidance across the five phases: Define, Measure, Analyze, Improve, and Control (DMAIC). It features three complete, built-in case studies and supports project selection, data collection planning, selection of statistical tools, hypothesis testing, Design of Experiments (DOE), improvement plan evaluation, the formulation of Statistical Process Control (SPC) plans, as well as interactive coaching and case-based instruction. This article introduces its positioning, scope of coverage, activation methods, typical phase-by-phase usage, and applicable scenarios. It serves as a valuable reference for professionals involved in quality, process engineering, and project management.
Read MoreDouyin Crawler: Searching for Viral Videos by Keyword
🚀 Douyin Crawler is a community-published data-analysis Skill, located at @user_fb71c45d/douyin-pachong, version 1.0.0. It searches for popular viral works on Douyin based on keywords, supports filtering by date range, and displays results in structured tables for finding Douyin hot content, viral videos, and work data. Official trigger words provided: Douyin work query, Douyin viral, Douyin hot, Douyin hot list, Douyin search, viral videos, hot videos. The tag requires_api_key=true indicates that an API key is required for operation; specific configuration fields have not been verified and will not be fabricated. Installation commands could not be opened and verified from the official page this time, and the captured materials were also not provided, so unverified commands are not listed; refer to the official page. Suitable for scenarios such as content topic selection, hot spot tracking, and viral sample collection.
Read MoreLean Cost Savings Calculator: Calculating the Benefits of Lean Improvement Projects Accurately
This article introduces the "Lean Cost Saving Calculator" Skill. Designed for benefit calculation scenarios in Lean Six Sigma and Continuous Improvement projects, it supports eight calculation methods: efficiency improvement, scrap reduction, variable cost reduction, freight reduction, inventory reduction, lead time reduction, layout optimization, and headcount optimization. It distinguishes between "Hard Savings" (recognizable in EBIT) and "Soft Savings" (capacity indicators), supports annualized calculations and incentive claim computations, and generates interactive HTML calculation reports with one click. The article outlines the Skill's positioning and origin (Community, published by user_a549da14, version 1.0.0, categorized as Professional, no API Key required), core features, typical workflows, as well as applicable scenarios and precautions regarding calculation standards. The official link is https://api.skillhub.cn/user_a549da14/ibp-003.
Read Moredsh-agents-visualization: Centralized Management of AGENTS.md and Multi-Tool Rules in DSH Settings
The DSH plugin dsh-agents-visualization maintained by LINGTIAN303 provides unified management of project-level directory chains, global ~/.dsh and ~/.agents, and user-level rule files for 25+ third-party AI tools within the "Rules" settings. It supports project discovery consistent with dsh-agent-instructions, Markdown preview and in-place editing, injected byte and token estimation, conflict and cross-tool duplication detection. Custom tool configurations are written to ~/.dsh/agents-visualization.json. Installation: dsh plugin --profile <profile> add dsh-agents-visualization. Requires DSH 0.2.x and Node >= 20, with version 0.1.2+ recommended. MIT license.
Read MoreCampus Psychological Crisis Intervention Process: Translating Plans into an Actionable Tiered Response
The crux of handling psychological crises on campus lies in whether contingency plans are truly put to use by frontline staff. This article introduces the Skill “Campus Psychological Crisis Intervention Process,” released by Kunlun Growth. Targeting homeroom teachers, school counselors, and moral-education administrators, the Skill covers student psychological risk signal identification, tiered crisis-intervention response, home-school communication, and referral workflows, aligning with all-staff heart-cultivating and mental-health promotion initiatives. Available at @kunlungrowth/edu-psych-crisis, Version 1.0.0, category: education; corporate source; no API key required. The article describes its positioning, four core capabilities, installation and activation procedures, and usage considerations, while clarifying that it cannot substitute for professional medical judgment.
Read Moredsh-dntb: DSH Model Inference Plugin on SkillHub
Introduce the DSH plugin dsh-dntb maintained by lingrenjie7-star. SkillHub lists it as installable, categorized under model inference, version v1.0.1. At the time of writing, the GitHub repository and README were not accessible. The main text contains only verified metadata and provides an installation prompt on the directory page. It also states that the plugin runs with dsh process privileges, and the source code and license must be reviewed before installation.
Read Moremonograph-writing: Breaking down long manuscripts into AI-relay-able and verifiable batch pipelines
monograph-writing is a community-released, general-purpose SKILL.md format Skill (@user_a72e0254, version 1.7.0, categorized under content-creation, requires_api_key=false) that transforms long-form manuscript writing into a batch-based pipeline, applicable to academic monographs, textbooks, technical books, and academic dissertations. Centered around six core mechanisms—academic red lines, four-category output classification, ledger-as-state, automated validation, data and figure red lines, and final manuscript polishing—it resolves two major pain points in AI writing: the propensity to hallucinate literature and data, and the loss of progress during cross-session continuation. Progress, figure and table numbering, verified bibliographies, data freezes, and figure/table evidence cards are all persisted to disk. Additionally, citation numbering, word count grading, and key numerical consistency are strictly audited by scripts, ensuring that a batch is only finalized after zero errors remain. This article details its features, installation and activation methods, typical use cases, and applicable limitations.
Read MoreBingjian: An Agent Skill Based on Traditional Wisdom of Character Assessment
Bingjian (Ice Mirror) is an Agent Skill focused on personnel assessment and social interaction, version 1.0.1, maintained by community publisher user_fd308b11. Categorized under SkillHub's "life-service" section, it originates from the community. It combines observational frameworks from traditional texts such as *Bingjian* and *Guiguzi* with concepts from Western psychology, game theory, behavioral economics, and neuroscience. It targets six specific tasks: analyzing personality traits, predicting behavioral trends, formulating interpersonal strategies, selecting communication tactics, managing conflict scenarios, and enhancing social insight. The SkillHub page lists a rating of 100,000, with 902 downloads, 0 installations, and a tag of `requires_api_key=false`. Installation generally follows the standard SkillHub SKILL.md conventions; however, the official page was inaccessible at the time of writing, and specific instructions for Cursor, Codex CLI, or Claude Code were not provided. Users should note that output quality depends on the input provided, and this tool does not replace professional judgment in legal, medical, or psychological fields.
Read MoreProgramming and Development Prompts Collection: 20 Scenarios Prompt Package
This is a community skill released on SkillHub by user_2c08c7ce. It is categorized under dev-programming and packages 20 prompts tailored to the programming and software development domain, designed to cover scenarios such as code review, refactoring suggestions, test case generation, technical documentation, and architecture design. This article introduces its positioning, coverage capabilities, verifiable configuration details, typical usage, and noteworthy considerations, including its pay-per-use model (at ¥9.90), community origins, and zero installations, along with its official link.
Read MoreAI Comic Drama Production Expert.Skill: Converting Novels into Directly Feedable Storyboard Prompts for AI Video Tools
**AI Comic Drama Production Expert.Skill** is a community skill published on SkillHub (Version 9.4.0, Category: design-media, Rating: 100000, Downloads: 984, No API Key required) at the coordinate @user_757bce4c/ai-comic-drama-studio. It consolidates the repetitive labor between novels and AI videos into structured deliverables, covering script adaptation, episode pacing, storyboards, character consistency locking, shot-by-shot prompts, voice-overs, subtitles, and publishing packaging. Users trigger the skill by pasting a novel excerpt or mentioning keywords such as comic drama, AI short dramas, storyboards, or character inconsistency. The output is a complete storyboard prompt list ready for AI video tools like Jimeng, Kling, vidu, and Runway. The official page has not disclosed verifiable installation commands or configuration fields; access is based on the coordinate reference, and it is recommended to verify details on the official page before installation. It is suitable for AI short dramas, anime commentary, novel promotion, and comic drama studios.
Read Morewind-alice: A CLI tool for invoking the Wind Alice Agent to execute specified Skills
Wind-Alice is a CLI tool developed by Wind Information. It invokes the Wind Alice Agent via the A2A protocol, executing specified sub-Skills in an SSE streaming manner and returning analysis results. This solves the issues of manual A2A client implementation and stream parsing for callers. It is suitable for scenarios requiring the specific invocation of Alice sub-Skills, such as investment research survey checklists, one-page investment memos, and financial information verification. The tag `requires_api_key=true` indicates that a key must be configured before use; official installation commands and interface fields should refer to the publisher's page.
Read MoreDeconstructing Viral Videos: Direct Generation of a Three-Format Breakdown Report from a Single E-commerce Feed Video
The "Viral Video Decode" Skill is a video analysis tool designed for e-commerce information flow scenarios. Located at `@user_4f12e6bd/bao-kuan-video-decode`, it was published by `user_4f12e6bd` (Version 1.0.1, Community Source, Category: business-ops). By inputting a single video, the Skill outputs a complete breakdown report in three formats: JSON, Markdown, and HTML. The report includes viral point identification, content direction, narrative structure, framework matching (Five-Act, PAS, AIDA, PAIPS, DR, Twist, Checklist, Pyramid), reusable scripts, compliance pitfall avoidance, a viral index score with optimization suggestions, and built-in compliance safeguards for beauty and oral efficacy categories. It features a built-in local agent, `scripts/video_ai.py`, responsible for frame extraction, multi-VLM analysis, and report generation. It can be launched via a drag-and-drop web interface or a single command. The developer claims the automatic workflow increases efficiency by approximately 6-8x for individual videos and ensures 100% standardized output. For a monthly volume of 200 videos, it saves approximately 127 hours (valued at roughly ¥10,000), with VLM costs ranging from ¥0.02-0.1 per video. This Skill is tagged with `requires_api_key` and can be installed and used within WorkBuddy without requiring manual configuration of VLM video language models. Source page data indicates a rating of 100,000 and 920 downloads, but 0 installations. Due to the extremely low public adoption rate, its reliability requires independent verification via small-sample testing.
Read MoreXiaoming's Selected Resale Property Cases: Delegating Solution Generation to Agents
Second-hand Housing Transaction Solution Optimizer (@user_256d1ea7/ershoufang, v1.0.2, community/professional) Solidifies the generation of second-hand housing solutions into Agent-triggerable paths: The first level targets first-time homebuyers who have not yet selected a property, outputting comparisons of multiple candidate property names and unit layout solutions; the second level, after the user locks in a property, outputs detailed solutions regarding the entire transaction process, risk prevention, and transfer with existing mortgage, and generates a compact, printable Word document. The introduction lists trigger words such as second-hand housing transaction, where to buy, property selection, second-hand housing process, transfer with existing mortgage, second-hand housing evaluation, and second-hand housing taxes, and clearly states it is not applicable for new property purchases, renting, commercial real estate transactions, or real estate legal litigation representation. Public tag requires_api_key=false, SkillHub rating 100000, downloads 885, installs 0. The provided materials do not include installation commands or differences between hosts; refer to the official SKILL.md.
Read MoreVideo Cover Generation: One-Step Cover Copy and 3:4 Image Prompts
Introducing the community skill “WeChat Channels Cover Generator” (@user_92725383/video-cover-generator, version 1.0.0, category: content-creation). It breaks down a Channels cover into two deliverables provided together: the cover copy and an AI image prompt that describes a 3:4 portrait composition. Its tag requires_api_key=false means no API key is required upfront. It does not generate the actual image, nor does it write scripts. The official page does not provide installation commands or configuration fields, and fetching the page body is unavailable, so for installation we only state that you should place it in the directory via your agent’s skill loading method and then reload; do not copy third‑party directory paths. For usage, we provide only the high‑level directions consistent with the brief (provide a topic, get the copy, get the image prompt, and manually align the margins), and note that “3:4 fit” refers to the prompt aspect ratio rather than bitmap dimensions; ratings and download counts are as of the time the page was opened.
Read MoreZodiac Service: Turning Zodiac Query and Horoscope Analysis into MCP Tools
The "Constellation Service" is a constellation-related capability package delivered via MCP (Model Context Protocol). Released by Xiao Ben Yang Technology, it is located at @org-rn88lg3j/constellation-svr with version 1.0.0, categorized as "life-service," and tagged with "requires_api_key=true." It exposes three types of tools to the host program: constellation information queries, horoscope analysis, and compatibility tests. This allows constellation-related Q&A within conversational applications to be transformed into callable and traceable tool calls, rather than relying on the model's memory. The available materials do not provide official installation commands, tool input parameters, or API Key configuration fields; this article describes the situation factually and does not elaborate on integration differences across Cursor, Codex CLI, and Claude Code. Before integration, please visit the official address to confirm the tool list, request format, and key configuration.
Read MoreWork Safety Experts: Turning GB/AQ/TSG Standard Numbers into Verifiable Compliance Answers
Work Safety Expert is a knowledge-based Agent Skill designed for the work safety domain, published by community publisher user_d7bf28e7, with source marked as community, categorized under knowledge-management, version 1.0.0, and organized according to the standard SKILL.md format. Its capabilities span 108 laws and regulations and 7,047 standards, answering work safety questions related to hidden hazard identification, compliance obligations, special operations, equipment acceptance, incident handling, personal protection, and emergency management. It is triggered when users inquire about the above scenarios or mention standard numbers such as GB, AQ, TSG, JGJ, or laws like the Work Safety Law and Fire Protection Law. Tagged with requires_api_key=false, it requires no API key configuration for use. Official documentation does not specify concrete installation commands or host differences; actual integration depends on the SKILL.md loading rules of the Agent host in use. The article also outlines key considerations including applicable user groups, retrieval scope limitations, standard currency, and version verification.
Read Moredsh-cot-translation-reader: Display Official Chain of Thought Machine Translation on the Reading Page
The official COT translation bundle places translated text under conversation.chat.reasoning.body, but the dsh-better-display reading view does not render this slot. Maintained by LingMoe404, this plugin depends on better-display ≥0.3.9 and the official experimental-cot-translation-bundle. It automatically translates reasoning blocks via the text-transform seam, and supports toggling between original and translated text on a per-card basis (requires better-display PR #82). It only modifies display and does not alter conversation history. To install: pnpm add dsh-cot-translation-reader, then restart DSH and add it to your profile bundles.
Read MorePICK-ME IP Location: An Automated Location Self-Check System
PICK-ME IP Positioning, created by @indiv-cunzhanglaoxie, is version 1.2.21, categorized under content-creation, and requires no API Key. It breaks down personal IP positioning into six key components: P (Product & Audience), I (Innovation/Differentiation), C (Communication/The Three Cs), K (Keywords/Mental Vocabulary), M (Investment/Commitment), and E (Evaluation/Closed-loop). Starting from three layers of authentic storytelling, five sources of trust, and verbatim audience quotes, it progresses through four white space opportunities, a dual-word strategy, and the A×B×C positioning formula, concluding with a layman's paraphrasing test. The output consists of a six-chapter diagnostic report and a one-page positioning card. Accompanying tools include a 40-minute self-assessment checklist, a 30-question interview guide, 30 topic generation methods, and a checklist of 12 positioning mistakes. It is suitable for creators and consultants preparing to launch or those struggling with long-term positioning, but not for those looking to simply copy a slogan.
Read Moredsh-life-hair: Automatically Snapshotting Checkpoint Summaries During Context Compression
dsh-life-hair (Life-Saving Hair) is a memory-oriented DSH plugin maintained by LingDangZaiXiang. When DSH compresses the context, it generates a checkpoint summary and discards earlier details. At the moment compression occurs, this plugin writes the summary verbatim to ~/.dsh/sticky-notes/归档/压缩归档-<时间>.md and can optionally wake the session to add a sticky note. It does not trigger compression and does not change compression settings; minLength and deduplication reduce false triggers. Install with: dsh plugin --profile <profile> add LingDangZaiXiang/dsh-life-hair. Recommended to be used together with Meredith2328/dsh-sticky-note to view archives. MIT-licensed, version 0.1.0.
Read MoreCatching Limit-Up Stocks via Call Auction: Standardizing the Stock Screening Process During the 9:15–9:25 Session
"Limit-Up Stock Capture during Call Auction" is a Skill in the professional category published by community publisher user_c5785c13. Version 1.0.0 is located at @user_c5785c13/zzt, marked as community-sourced, with a rating of 100,000, 905 downloads, 0 installs, and a tag of requires_api_key=false. It organizes the daily stock selection workflow from the "Volatile Stock Seal Order Mystery Judgment Extension" course into actionable steps. It includes five capabilities: capturing snapshots at 9:15, 9:20, and 9:25, judging sentiment gates, locking the strongest themes, comparing the strengthening of thematic sentiment, and maintaining the web-based stock selection system. It is triggered when users request stock selection during call auctions, intraday limit-up capture, course-based stock selection, or limit-up capture via call auctions. This article introduces its positioning, core functions, official entry point, typical workflow, target audience, and notes regarding data sources, real-time performance, and risks.
Read MoreOne-Click Long Content Summarizer: Condense Web Pages, Subtitles, and Long PDFs into Structured Bullet Points
The "Long-Content One-Click Summarizer" (@user_e562897e/web-content-summarizer) is a community-sourced Skill categorized under knowledge-management on SkillHub. Version 0.1.1, it is published and maintained by user_e562897e. It compresses web pages, articles, video subtitles, and long PDFs into structured key points and action items. Based on the universal SKILL.md format, it does not require an additional API key as indicated by the publisher. The article outlines its capability boundaries, how to obtain and enable it from the official page, typical usage patterns, applicable scenarios, and precautions: the version is in its early stages with limited documentation, the output consists of key points rather than rewritten original text, and it cannot be used as a citation source. The ratings and download numbers cited in the text reflect the page values at the time of scraping.
Read More