K-12 Mathematics Teaching and Research Assistant: Introduction to a Skill for General School Mathematics Teaching and Learning
This article introduces the community Skill "K-12 Mathematics Teaching and Research Assistant." Designed for mathematics teaching and learning in mainstream primary and secondary schools across mainland China, it aligns with current curriculum standards and student grades to deliver concept explanations, standardized problem-solving, error diagnosis, tiered exercises, lesson plan design, and assessment feedback. It also generates offline interactive lesson pages upon specific request. This tool is suitable for lesson preparation by teachers, student tutoring, parental guidance, and the creation of question content for primary, middle, and high school mathematics. However, it is not intended for mathematics competitions or university-level mathematics, nor is it used to make assertions regarding high-stakes entrance exam policies (Zhongkao and Gaokao) without supporting local documentation. A snapshot of the official page indicates a rating of 100,000, 1,073 downloads, and `requires_api_key=false`. Due to a failure in fetching the official page, this article does not provide unverified installation commands; instead, it details only verifiable fields and provides the official URL for confirmation.
Read MoreWebsite Construction and Landing Page Design: Standardizing the Creation Process of Bright Tech-Style Landing Pages into a Skill
"Website Construction & Landing Page Design" is a Skill published by Tianshu Tanjie, located at @aborseek/build-landing-page, version 1.1.0, categorized under dev-programming, and does not rely on external API keys. It consolidates enterprise website development experience into four stages: design, construction, renovation, and review. It targets bright tech-style websites and responsive landing pages, suitable for enterprise, product, service, course, and event pages. This article covers its positioning, verified capability boundaries, installation and enablement methods based on the universal SKILL.md format, typical use cases and limitations, and notes that missing content in official documentation (such as specific templates, component lists, reproducible examples, and installation commands) has not been speculated upon. The Skill has a SkillHub rating of 100,000, 1,025 downloads, and 0 installations, with limited community validation.
Read Moredsh-session-delete: Providing True Session Deletion for DeepSeek Harness
The positioning of EndlessPoem’s dsh-session-delete on SkillHub and dsh.so is a client-side “true session deletion”—emphasizing erase rather than archive. This article is based on information that can be verified via directory introductions and community indexes. As the GitHub README could not be retrieved and the repository is currently inaccessible, installation commands and UI steps are not fabricated; instead, only the SkillHub installation prompt snippet and the installation spec recorded on dsh.so (github:EndlessPoem/dsh-session-delete) are cited. It also notes that the plugin operates on local data with the permissions of the dsh process, and that the source code and the MIT License should be read before installation.
Read MoreAcceptance Report Skill: Turning Delivery Acceptance into a Verifiable, Signable, and Traceable Checklist
The "Acceptance Report" is an Agent Skill released by Kunlun Growth (version 1.3.0), categorized under "office-efficiency" and requiring no API Key. It breaks down delivery acceptance into five steps: scope verification, deliverables checklist, criteria comparison, issue logging, and sign-off confirmation. It features built-in templates for three categories of acceptance criteria, an evidence retention checklist, a three-tier classification for issue closure, and a final payment risk assessment mechanism, covering software delivery, project completion, service procurement, and event content acceptance. This article introduces its capabilities, enabling methods, and typical use cases, while outlining its applicable scenarios and limitations.
Read MoreEnd-to-End SEO/GEO Workflow for Independent Websites: From Competitor Monitoring to Content Execution via a Single Entry Point
This is an SEO/GEO (Generative Engine Optimization) Agent Skill designed for independent websites and cross-border e-commerce sites. Published by user_50bad851, it originates from the community, is versioned at 1.0.0, classified under content-creation, and tagged with requires_api_key=false. It consolidates a complete closed-loop workflow into a single entry point: competitor website discovery (including visual reverse search and web-wide keyword search) → competitor price, promotion, and inventory monitoring dashboards → product keyword research, coverage, and gap queries → overall website SEO/GEO layout (information architecture, topic clusters, column planning) → phased and page-level optimization audits and technical SEO checks → optimized content implementation → structured data (JSON-LD) → llms.txt/robots.txt configuration and AI engine visibility → "de-AI-ified" blog writing. The author declares this to be the sole entry point for SEO/GEO content operations. It has merged the capabilities of the original seo-geo-optimizer; GSC/Ahrefs/Trends data pipelines are integrated with the seo-ops skill once credentials are configured. It is triggered by directly stating requirements regarding competitors, keywords, audits, schema, GEO, or blog writing, allowing the skill to determine the appropriate workflow stage. The figures in this text reflect data at the time of scraping: Rating 100,000, Downloads 1,014, Installs 0.
Read MoreJob AI Agent Prompt Generator: Converting Job Duty Manuals to 7-Layer AI Agent Prompts
Introducing the "Job Agent Prompt Generator" on SkillHub: it systematically transforms traditional enterprises' Job Description (JD) documents into system prompts for large-model-based digital employees (AI Agents). Based on Chu Daxia's seven-layer standardized prompt template for job-specific digital employees, it automatically retrieves JD documents from a knowledge base and performs in-depth parsing, producing a structured, seven-layer complete card architecture. Published by user_84457ab0; sourced from the community; category: ai-agent; version: 1.0.0; tag: requires_api_key=false (no API key required). Trigger scenarios cover JD documents, job responsibilities, digital employee prompts, AI Agent prompts, job agents, and job role prompt generation. Publicly available materials do not provide the specific names of the seven layers, the installation command, or reproducible examples. This article states this factually, makes no speculation, and provides applicable scenarios along with recommendations for validation before integration.
Read MoreLitigation Case Timeline Visualization: Organizing the Litigation Process into Interactive HTML
"Litigation Timeline Visualization" is a community-sourced Skill in the professional category, located at @user_281f816d/litigation-timeline, authored and released by Attorney Zhang Chenlu (version 1.0.0), and requires no API key. Designed for presenting litigation case proceedings, it provides three modes: a swimlane mode for comparing multiple parties in parallel or multiple procedures, a comparison mode for contrasting two sets of data (e.g., amounts payable vs. actually paid, claims vs. judgments), and a single-axis mode as a classic timeline. Supporting capabilities include intelligent document parsing and one-click generation of interactive HTML files, so users do not need to write any front-end code themselves. This article outlines a general workflow in four steps—preparing materials, selecting a mode, generating output, and manual review—and notes limitations: official materials only provide a feature overview and do not include installation commands, CLI directory conventions, or specific interaction details of the HTML; parsing results must be compared item by item against the original documents; time-sensitive data such as ratings and download counts should be based on the live display on the official page.
Read Moredsh-ip-info: View Public IP and Location in the Harness Interface
dsh-ip-info is a DSH networking tool plugin maintained by EndlessPoem, targeting the web profile. It can be installed via the command `dsh plugin --profile web add dsh-ip-info`. The plugin displays public IP, location, and carrier information either below the Composer or in a floating card. Its font size adapts to the Harness, and the display can be disabled. It is ideal for quickly verifying public network identity when debugging egress and proxy settings within the Web Harness. The plugin performs network queries using the dsh process permissions; please review the source code and license before installation. As the GitHub source may be unavailable, refer to npm and community directories.
Read MoreBusiness Development Team - Competency Model and Assessment for Distributor Management: Competency Modeling and Assessment of Pharmaceutical Business Development Teams
This article discusses a "proficient" skill published on SkillHub by the Baitong AI Service Platform, version 1.0.0. Designed for pharmaceutical industry investment promotion teams (focusing on agent management, investment agency, and channel deepening), it is based on the DDI Success Model methodology. The core content includes a three-level competency model PPT, tailored online assessments (a complete 200-question version and a simplified 100-question version), a 1,200-question Excel question bank with answers, downloadable personal reports, a PPT template for team reports, and a proposal for 20 systematic courses. The skill requires no API key (requires_api_key=false), has a SkillHub rating of 100,000, and has been downloaded 959 times. The article introduces its positioning, core capabilities, installation and enablement methods, typical usage via official trigger words, as well as applicable scenarios and limitations: the materials are biased towards pharmaceutical investment scenarios and require recalibration for cross-industry use, and the conclusions of the reports should not be directly used as a basis for personnel decisions.
Read MoreHow Dankoe Reshapes Your Life in a Day: Turning Life Re-engineering into a Daily Feedback Loop
# Preface When you want to change your life, the common approach is to set a grand goal: changing careers, losing weight, learning a new skill, or writing every day. There is nothing wrong with the goal itself; the problem is that it resides at the "intent" level. Within a few days, it gets swallowed by daily inertia, leaving only a sense of frustration. This skill treats life modification as an engineering problem: instead of relying on a one-time surge of emotion, it focuses on completing one minimal action every day, harvesting the results in the evening, and using feedback to correct the direction for the next day. The following sections introduce its positioning, capabilities, and usage. # What is this Dankoe How to Fix Your Entire Life in 1 Day is a Skill designed for self-improvement, located at `@user_d0deb03a/dankoe-how-to-fix-your-entire-life-in-1-day`, version
Read MoreFinancial Data Assistant: An Automated Skill for Repetitive Business Tasks in White-Collar Work
This article introduces the Finance Data Assistant ( @org-02qudk26/alpha-vantage-automation ) on SkillHub. The Skill is published by Zhiyu Technology, with the source listed as enterprise, the category as professional, and the version as 1.0.0. It targets business automation of repetitive tasks for office workers, producing reusable deliverables in a structured format, and is officially marked as requiring no API Key. The article summarizes its basic information, core capabilities, the enablement approach via coordinate-based import, typical usage, and applicable scenarios, and notes cautions such as limited public documentation, zero installations, and that it is an early-stage version, to help users evaluate it before adoption.
Read MoreMaterial and Energy Balances in Chemical Engineering: The Digital Foundation of Chemical Design
Chemical Material and Energy Balance Skill, coordinates @user_a8c636dd/material-energy-balance, version 1.0.0, community-released, Professional category. It is positioned as the "digital backbone" of chemical design, addressing three core issues: chemical identification and key property retrieval; stoichiometry of main and side reactions alongside whole-process material balance (in reactions, separation, and circulation); and calculation of reaction heat, phase transition heat, sensible heat, and utility loads. It outputs structured data for consumption by the equipment sizing Skill. It is based on the standards HG/T 20570-2014, GB/T 16483-2008, and the GB 30000 series (Globally Harmonized System of Hazardous Chemicals). The SkillHub page snapshot indicates a rating score of 100000, 1051 downloads, and 0 installs, with requires_api_key=false. No installation commands or reproducible examples have been provided in the official documentation; this limitation is honestly addressed in the main text.
Read MoreCapCut Rapid Production Assistant: The Shortest Route to Short Videos
The "Jianying Quick Creation Assistant" is a CapCut rapid output Skill released by the community. Currently at version 1.0.0, it falls under the SkillHub design-media category. With a `requires_api_key=false` tag, it has a rating of 100,000 and 1,077 downloads. It covers Text-to-Video, Image-to-Video, One-Click Creation, and Template Application, as well as individual functions like AI Digital Humans, Smart Captions, and Smart Cutout. It supports the mobile version of Jianying (CapCut), the Pro (Desktop) version, and the international CapCut. This Skill is triggered when users ask questions such as "How to edit quickly in Jianying," "How to produce videos in minutes," or "How to use Jianying features," providing the specific operational paths for the corresponding platform. The article outlines its positioning, core capabilities, coverage, installation source, and typical trigger scenarios. It also notes that the Skill only maps out operational pathways, while the actual editing is performed within Jianying, and highlights potential differences in interface access across different platforms.
Read MoreA-Share Sentiment Meter: Quantifying Market Sentiment via Six Dimensions to Provide Suggestions on Phased Position Building and Selling
A-Stock Sentiment Thermometer is a community Skill on SkillHub (Coordinates: @user_538520f4/a-stock-sentiment, Version 1.0.1, Category: professional). It quantifies A-share market sentiment into a 0-100 score based on six dimensions: 5-day momentum (15%), RSI(14) (15%), Volume Ratio (10%), 20-day historical volatility (20%), MA250 Deviation (20%), and Advance-Decline Ratio (20%). It provides market data and a dimension-by-dimension breakdown for four major indices: Shanghai Composite, Shenzhen Component, CSI 300, and ChiNext, ensuring transparent, non-black-box scoring. When sentiment enters the "Fear Zone," it generates signals for phased position building (triggered at scores of 34, 29, 25, and 21, with corresponding position allocations of 15%, 20%, 35%, and 30%). Upon reaching the "Greed Zone," it avoids mechanical selling by first evaluating four trend indicators: MA alignment, price standing above MA10, 5-day gain, and volume ratio. Each data dimension utilizes dual primary/backup sources from Tencent, Sina, and Eastmoney, with automatic failover in case of single-source failure. The official release includes seven rounds of backtesting based on the CSI 300 ETF (510300) from 2018 to 2026, showing an 85.7% win rate, while noting that users must adjust parameters for other assets. It supports scheduled daily execution at 9:25 with email push notifications. No API Key is required; trigger keywords include "A-share sentiment analysis," "market sentiment," "sentiment thermometer," and "fear & greed index." It is suitable for retail investors, contrarian investors, users practicing phased buying, and busy professionals, though signals rely on market data and scores serve as auxiliary references rather than guaranteed returns.
Read MoreLocal HTML Publication Tool: Instantly Convert Local HTML into Public URLs
The Local HTML Public Publisher (deploying local HTML to the cloud for universal access) is an Agent Skill based on the universal SKILL.md format. Coordinates: @user_af163eaa/z-html-public-deploy. Publisher: user_af163eaa. Version: 1.0.1. Category: dev-programming. It encapsulates CloudStudio static site deployment, allowing one-click deployment of local self-contained HTML files or folders containing index.html into publicly accessible links, returning a public URL and previewing it within the conversation. It is triggered when users say "publish this html / turn into a link / deploy / share link / publish html / generate access address / online preview". Tag: requires_api_key=false. According to the SkillHub page, the rating is 100000, downloads are 980, and installs are 0.
Read MoreFinancial Early Warning Skill: A-Share Three Statements Anomaly Signal Scanning and Multi-Source Cross-Validation without API Key
A financial early warning engine for A-shares, scanning nine types of financial anomaly signals across multiple periods of the Three Major Financial Statements. It triggers three-level alerts (🟡🟠🔴) after confirmation across three consecutive periods, outputting alert_list.csv, alert_detail.md, and safe_list.csv. Uses westock-data as the primary source (free and requires no API key), with HTTP cross-validation via Eastmoney and optional enhancement using iWenCai. This is a pure early warning system and does not include valuation or stock selection logic.
Read MoreBaidu Netdisk Duplicate File Cleanup: Scanning, List Confirmation, and Batch Deletion
The Baidu Netdisk web version lacks built-in deduplication capabilities, requiring manual sorting and comparison to identify duplicate files with the same name and size, followed by individual selection for deletion. This community-maintained Agent Skill (@user_db3a3bce/baidu-netdisk-dedup, Version 1.1.0, Category: ai-agent) automates this workflow. It utilizes the agent-browser backend to log into the Baidu Netdisk web interface, recursively scans all files via the in-page /api/list interface, and identifies duplicates based on "filename and size." It generates a list for user confirmation before executing batch deletions to the recycle bin via actual UI interactions (Playwright clicks). Trigger phrases include "clean up netdisk duplicate files," "netdisk deduplication," "Baidu Netdisk slimming down," and "delete duplicate materials in netdisk." It will not trigger for non-Baidu Netdisk operations or during simple upload, download, or sharing tasks. The tag requires_api_key=false indicates that no additional API Key is needed. This article outlines applicable scenarios and precautions: the identification rule relies solely on filename and size, and deletions are moved to the recycle bin rather than being permanent.
Read More1TodoS To-Do List: Connecting AI Assistants to Your Todo System
1TodoS To-do List is a community-sourced Skill located at @user_9c85a73f/kairusi-1todos. Categorized under life-service and maintained by user_9c85a73f (Version 1.0.2), it requires an API Key and prior authorization of a 1TodoS account. It enables AI assistants that support Skills—such as Claude Code, Cursor, WorkBuddy, and OpenClaw—to connect to your 1TodoS account for natural language to-do management. Verified capabilities include: viewing all projects and lists, retrieving pending or completed tasks by list, global keyword search, creating new tasks in specified lists or the Inbox, marking tasks as complete, and deleting tasks. The article details prerequisites, typical usage (one-sentence creation, viewing, searching, and updating), target audiences, and caveats regarding its community source.
Read Moredsh-llm-pi-ai-image-offload: Restore image token metering for pi-ai routing
In DSH, the pi-ai route leaves imageRequestPricing unimplemented by default, so dsh-spill-policy cannot estimate the tokens of oversized base64 images in tool results, causing inline content to overflow the context window. The community plugin dsh-llm-pi-ai-image-offload restores a byte-based pricing calculator on adapter instances and, together with maxInlineTokens, triggers offload; models can still read the images back via read_image. Based on the official README, this article explains the root cause of the issue, the pricing formula, optional configuration, GitHub installation commands and bundle activation steps, and points out debug verification methods along with extension point risks.
Read MoreMaster of Live Commerce Scripts: Instantly Generate 5 Script Categories, Featuring Loop Structures and a Banned Phrases Reference
Livestream e-commerce teams often face two problems: on-the-fly scripting by hosts and traffic throttling triggered by prohibited phrases. Douease’s Livestream Sales Script Master targets this scenario: input product information and the average order value, and it outputs five script types—opening hooks for audience retention, delayed offers, urgency-closing prompts, smooth product pivots, and direct calls-to-action to order—each with templates and examples. A built-in 5-minute looping script structure keeps new viewers who arrive at any time engaged. It offers playbooks for three price tiers: 9.9 traffic-driver SKUs, profit-focused products around one hundred yuan, and high-ticket items. It also includes a compliance glossary of prohibited livestream terms (e.g., “lowest price,” “number one across the entire web”) with approved replacements, as well as a rundown (schedule) for pacing and product sequencing throughout the session. No API key is required, and it covers Douyin, Kuaishou, WeChat Channels, and Taobao Live. This article introduces its positioning, core features, how to install and enable it, typical use cases, and usage considerations. It also notes that no specific installation commands are provided in the materials; please verify on the official page.
Read MoreSystem Slimmer: Securely Cleaning Disk Junk with Agents
The System Slimmer (@user_af163eaa/z-system-slimmer) is an it-ops-security skill published by the SkillHub community, version 1.0.0. It scans caches, logs, temporary files, leftover uninstall files, and duplicate files, and generates a three-tier (green/yellow/red) cleanup checklist. Upon user confirmation, files are deleted via the recycle bin. It only clears categories that can be safely deleted, without touching system core files or user documents. The "Allow Full Access" permission must be enabled on the agent platform for it to actually execute file changes. It is activated via trigger phrases such as "system slimming," "clean up junk," "free up space," and "shrink the C drive."
Read MoreOrganization and Personnel Newspaper Writing Skill: Formalizing Newspaper Submission Guidelines into Callable Processes
This is a Chinese writing skill designed for submissions to the *China Organization and Personnel News*. Published by user_4c93aead on SkillHub as a community contribution, it is categorized under content creation with the identifier @user_4c93aead/zuzhirenshiwriter (Version 1.0.1). Its core capability involves generating articles that meet newspaper publication standards by combining user prompts with web-searched materials. It covers topical areas including Party building, talent management, cadre affairs, grassroots Party building, and commentary on organizational work. The documentation outlines its positioning, core functions, installation and activation entry points, and usage workflows. It also notes that official installation commands are absent from the provided materials, ratings and download figures are for reference only, community-sourced skills lack official endorsement, and human verification of facts and citations remains necessary for final drafts.
Read Moredsh-handoff-doc: Hand off long sessions into a HANDOFF.md
The DSH plugin dsh-handoff-doc (ElioFanyu, v0.2.0) is the successor to dsh-session-handoff: it rewrites HANDOFF.md with a validity period in full at Step boundaries, archives the previous version to CHANGELOG and history, and sends only a short pointer to read the document to the new session. This section introduces the ten-section document skeleton, the /handoff command, installation and auto-trigger configuration, as well as the design and limitations related to uninterrupted execution and workspace write validation.
Read MoreAuditing Shrimp 3: Pre-Installation Security Check for AI Agent Skills
Audit Shrimp 3 (@user_d93c4aed/qmi-skill-vetter, version 1.0.1) is a community-sourced skill on SkillHub, serving as a security auditing tool for AI Agent skills. Its purpose is to shift inspection actions to a pre-installation phase: prior to installing skills from sources such as ClawdHub or GitHub, it scans for red flags, checks for overreach in permission scopes, and identifies suspicious calling patterns such as silent external communications or unconstrained execution. It requires no API key itself. The article details its positioning, four verifiable core capabilities, and a minimal usage workflow, while also noting information gaps identified during the review: as the official page could not be scraped successfully, installation commands and configuration fields are not provided. Furthermore, it notes that this is a community-sourced skill with zero installations, emphasizing that its audit output is not a substitute for manual code review and runtime permission isolation.
Read MoreSkill Coach: Defining Direction and Conducting Reviews for AI Skills
Skill Coach is an AI-agent skill released by community developer user_dc515faf. Version 1.0.1, it is based on the universal SKILL.md format and requires no API key. It offers two modes centered on AI skills themselves: Design Mode and Review Mode. Design Mode is used when users want to create a new skill or ask how to design one. It provides design direction using the reverse R-V-C-E (Risk → Value → Craft → Evidence) mental model, covering risk prediction, existence validation, trigger boundary definition, determinism strategies, and verification methods. Review Mode is used to evaluate existing skills, optimize descriptions or trigger conditions, determine go-live or merge viability, and analyze trigger conflicts between skills. The skill is activated by Chinese and English trigger words, such as "create a skill," "new skill," "review skill," "optimize skill," "skill quality," "skill go-live decision," "inaccurate skill trigger," and "skill trigger conflict." If the objective is unclear, it is recommended to use "grill" first to clarify requirements before providing design direction. It is explicitly not intended for general coding debugging unrelated to skills, product process design unrelated to skills, or general Q&A that does not concern how to create or evaluate a specific skill. SkillHub Rating: 100000; Downloads: 1120.
Read More