Sif-Amazon-Research: Amazon Research and Diagnosis Skill Based on Real Sif Data
sif-amazon-research is an Amazon e-commerce research Agent Skill published on SkillHub that integrates with the Sif MCP server. Leveraging Sif's authentic data on market trends, traffic, keywords, sales, and advertising, it facilitates market validation, competitor analysis, traffic root cause diagnostics, keyword opportunity discovery, advertising structure reviews, listing assessments, and growth optimization. It triggers upon user mentions of Sif, ASIN research, traffic decline, competitor analysis, ad reviews, or inquiries regarding market entry, expansion, or exit. As the official page does not provide installation commands or client compatibility details, verification is required prior to integration.
Read MoreHome-Buying Decision Assistant: Breaking House-Viewing Decisions into Verifiable Dimensions
Home-Buying Decision Assistant (handle: @user_af163eaa/z-home-buy-decision) is a Skill listed under SkillHub’s life-service category. Version 1.0.1. Publisher: user_af163eaa. Source: community. No API key required. It breaks home-buying decisions into fixed dimensions—location, amenities, layout, orientation, floor level, developer, property management, total price, monthly payments—and provides an analytical framework. It compares new-build vs. resale homes and flags pitfalls such as common area sharing, parking spaces, and school-district promises. It is positioned to offer structured analysis rather than purchase conclusions. It explicitly states that contracts and local policies prevail and does not replace professional assessments. This article introduces its core features, how to enable it based on the standard SKILL.md, and its applicable scenarios and limitations.
Read MoreOvernight Stock Selector Pro: End-of-Day A-Share Overnight Arbitrage Screening Skill
This article introduces the community-sourced A-share overnight stock screening Skill, “Overnight Stock Screener Pro” (ID: @user_b0160c90/overnight-stock-screener, v1.0.0, source: community, category: professional, published and maintained by user_b0160c90, tag: requires_api_key=false). After 14:30 in the late trading session, it scans A-shares and applies a combined filter based on Yang Yongxing’s overnight arbitrage strategy and the “one-night hold” method with eight screening criteria. For each candidate stock, it outputs an overnight win-rate score (a 0–100 thermometer), main capital flow, hot-money seats on the Dragon and Tiger List, four sell-response playbooks for the next day’s open, and market-timing risk controls. It can be triggered by keywords such as overnight stock screening, late-session stock selection, one-night holding, what to buy tomorrow, late-session limit-up scanning, the overnight method, and Yang Yongxing’s strategy. Official URL: https://api.skillhub.cn/user_b0160c90/overnight-stock-screener; SkillHub page: https://www.skillhub.cn/skills/user_b0160c90/overnight-stock-screener. The article also notes that the official source does not provide verifiable installation commands or configuration fields, that it must be enabled according to the SKILL.md runtime conventions, and that time-sensitivity applies, it does not constitute investment advice, and data sources should be verified independently.
Read Moredsh-masterduel: Search Cards, Build Decks, and Inspect Master Duel Decks in DSH
dsh-masterduel is a DSH client plugin maintained by LOESS16 (0.3.0, MIT). Locally, it exposes /dsh-masterduel/panel and a REST API, and via the master-duel-ai-advisor MCP enables card lookup, AI-based effect search, related-card queries, three-column deck building (collection, drag-and-drop, EDOPro first-turn testing, hand_odds probabilities, YDK/YDKE), rankings, and deck check. It requires DSH Desktop 0.2.x, Node 22.13+, and an MCP deployment. Installation follows the README: manual copying or dsh plugin add with an absolute path. This article summarizes features, dependencies, and typical API usage based on the official README, and outlines permission and network considerations.
Read MorePrecision Customer Acquisition via Short Video: Sales-Driven Scripts for Local Businesses and Knowledge Commerce
This article introduces the community Skill "Short Video Precision Lead Generation" (@user_423c8d0a/short-video-lead-generation, v1.4.0, content-creation). It breaks down short video customer acquisition into seven capabilities: acquisition scripts, voiceover copy, matrix topic selection, viral video reshoot analysis, comment and DM handling, livestream and local promotion materials, and a 72-hour validation plan. Targeting seven business types—Chinese local merchants, same-city stores, e-commerce, paid knowledge, AI service packages, personal IPs, and B2B services—its goal is closed deals rather than generic traffic. The article provides the scope of trigger words, configuration instructions without needing an API Key, official addresses, and three reproducible demand examples. It also notes caveats such as the installation command being unverified, the 72-hour plan not being a guarantee of results, and the need to comply with platform and regulatory requirements during the lead handling process.
Read Moredsh-bing-wallpaper: Use Bing Daily Wallpaper as DSH Interface Background
dsh-bing-wallpaper is a DSH wallpaper plugin maintained by lmxcode that displays the Bing daily wallpaper as the full background of the Web interface. It supports automatic wallpaper updates at midnight and upon returning to the tab, allows pinning a specific day within the last 30 days, and relies on Host and localStorage caching when offline. The installation command is `dsh plugin --profile <profile> add dsh-bing-wallpaper`, with the optional `quality: uhd` parameter to enable 4K resolution. The operation entry is located at the far left of the info bar below the conversation input box.
Read MoreWeChat Channels New Account Script Generator: Get a Ready-to-Record 60-90s Voiceover Script Instantly
This Community Skill is designed for novice WeChat Channels creators. By inputting a content niche, account positioning, and specific pain points, it generates ready-to-record 60-to-90-second spoken scripts. The output includes a "Golden 3-Second" hook (4 customizable types), practical advice on avoiding common pitfalls (covering 8 misconceptions, revealing the truth and providing solutions), essential knowledge for beginners, precise topic hashtags, publishing configuration settings, and a compliance self-check checklist. The entire process is altruistic, purely practical, and devoid of empty rhetoric. All statements regarding algorithms and traffic are labeled as operational experience references; no promises of follower growth or monetization are made. It incorporates built-in reminders regarding platform compliance red lines and AI content labeling. API Key requirement: false. As installation commands and editor-specific differences were not found in the retrieved materials, the text honestly states that details must be confirmed on the official page. This tool is ideal for new creators in the cold start phase.
Read Moreyq-video-motion-analyzer: Decomposing instructional videos into keyframes, pose analysis, and stick figure diagrams
yq-video-motion-analyzer is a community Skill located at @user_15292d5a/yjkj-yq-video-motion-analyzer. Version 1.0.0, categorized under design-media, published by user_15292d5a, and sourced from the community. Targeting sports instructional videos, it integrates keyframe extraction, motion posture analysis, stick figure motion diagrams, and improvement suggestions, reducing the manual effort of scrubbing progress bars, taking screenshots, and relying on subjective judgment. Official page data shows a rating of 100000, 887 downloads, and 0 installations. The tag `requires_api_key=false` indicates that the Skill itself does not require an additional API key, but its actual operation still depends on the models and video understanding capabilities behind the client. The article introduces its capability boundaries, activation methods based on SKILL.md, typical usage organized by trigger keywords, as well as applicable scenarios and limitations.
Read MoreMultisim-toolkit: Enabling DSH to Drive Local NI Multisim 14.3
multisim-toolkit (dsh-multisim-toolkit 0.1.0) is a DSH client plugin maintained by llyy-t. It connects Multisim 14.3 COM to the GUI via a 32-bit Python worker, registering nine tools: msim_status, circuit, inspect, edit, analyze, export, gui, spice, and design. It supports opening and closing .ms14 files, parameter modification, DC/AC/transient and SPICE analyses, exporting netlists/BOMs/schematics, and taking screenshots of real circuits and waveform windows. The same repository contains skill and reference. The installation command is `dsh plugin --profile <profile> add github:llyy-t/multisim-toolkit`. It requires Windows, Multisim 14.3, and 32-bit Python with pywin32. MIT licensed.
Read Moredsh-live-preview: Live Preview Your Application While Coding in DSH
**dsh-live-preview**, a DSH client plugin maintained by **llr-sketch**, provides real-time in-window previews of web dev servers, Windows desktop programs, and Android physical devices/emulators (30–60fps mirroring). Its panel allows one-click installation of the Android emulator and dependencies, while displaying process logs. It supports importing APK/EXE/HTML files, version and history management, and enables the Agent to automatically launch apps via *live_preview*. Install using `plugin_manager install_bundle`, with support for local directories, npm@0.2.4, and Releases. It is ideal for interview acceptance, teaching demonstrations, and AI self-verification; full Android capabilities require Windows 10/11 x64 and virtualization.
Read MoreHiQ Cortex — LCA Data Query: Replacing Recollection with Real Emission Factors
HiQ Cortex — LCA Data Query is a community-published professional Skill designed to replace models reciting numbers from memory when real emission factors are needed. It covers 18 life cycle inventory databases (Ecoinvent, BAFU, USLCI, ELCD, EF, worldsteel, AusLCI, HiQLCD, etc.) and 24,000+ published EPDs. It supports tasks such as material GWP queries, product carbon footprint calculations, BOM carbon accounting, industry benchmarking and percentile positioning, production route comparisons (BOF vs. EAF steel, primary vs. recycled aluminum, grey vs. green hydrogen), and EPD peer review. An API Key is required. The article presents its positioning, core capabilities, activation methods, typical use cases, and usage considerations.
Read MoreAutoHOT Automotive Daily Intelligence: A Skill for Daily Automotive Industry Intelligence Retrieval and Briefing Generation Based on Public APIs
The "AutoHOT Daily Automotive Intelligence" is a general-purpose Skill in SKILL.md format published on SkillHub. It is located at @user_cae83fba/autohot-skill, published by user_cae83fba, with a Source of Community, Category of Professional, and current version 2.2.0. It queries four types of content—automotive hotspots, daily reports, brand dynamics, and industry track trends—via the public AutoHOT API, and generates three types of deliverables: sales morning briefings, executive weekly reports, and event impact analyses. It is designed to replace the repetitive manual work of sifting through information and compiling briefings. The page tag `contains requires_api_key=true` indicates that an API Key must be configured for use. As of November 5, 2026, the page statistics show 100,000 ratings, 991 downloads, and 0 installations. The article covers the Skill's positioning, core features, installation and activation methods (including the official address and coordinates), typical use cases, applicable scenarios, and limitations (API Key requirement, community-sourced nature, dynamic data, and model-dependent output). It also notes that the fetched materials do not include specific installation commands, so users should refer to the official page for accurate instructions.
Read Moredsh-token-meter: View Bailian Token Plan usage in the DSH sidebar
dsh-token-meter (package name: dsh-plan-meter) is a DSH client plugin maintained by lkpsg. It leverages the official bailian-cli login session to display the monthly/weekly/5-hour quota, subscription, credits, and reset countdown of Bailian Token Plan in the Web UI, supporting installation for both Desktop and web profiles. This article, based on the repository README, describes the dependencies, installation commands, configuration options, and usage limitations, and does not cover Cookie or API Key solutions.
Read MoreKuaishou Content Search: Retrieve works via keywords and obtain interaction data instantly
Kuaishou Work Search (@user_5f9c21aa/kuaishou-search-redfox) is a community-sourced Skill classified under content creation, version 1.0.0, published by user_5f9c21aa. Enter a keyword to search for popular Kuaishou works, with filters for sort order and publish time. Returns work title, author, play count, like count, comment count, and collect count, and provides a detail-page link. Trigger phrases include Kuaishou search, Kuaishou works, Kuaishou trending, Kuaishou video query, and Search Kuaishou. Metadata tag: requires_api_key=true; an API key must be configured before use. The official page could not be retrieved this time; installation commands and configuration fields could not be verified, and no unverified commands are listed in the article. The article also describes typical usage, applicable scenarios, and precautions.
Read Moredsh-formula-verify: Replacing Oral Proofs in Dialogue with Executable Verification
dsh-formula-verify is a DeepSeek Harness community plugin maintained by lizyoko9, providing formula verification skills, a bootstrapped SymPy CLI, and a Formula Lab sidebar panel. The installation source is `git+https://github.com/lizyoko9/dsh-formula-verify`. Core capabilities include symbolic/numeric/failed validation checks, pytest execution within a formula card and private venv, paper extraction from TeX/PDF/Markdown, and client-side conversion and extraction workflows. It is suitable for reading papers and documented verification; however, please note that SymPy does not constitute formal proof, and consider the trust boundaries regarding PDF/Mathpix and external commands. Please review the MIT source code before installation.
Read MoreInvestment Due Diligence Officer: A Due Diligence Operating System for Translating Interview Findings into Valuation Discounts and Terms
Produced by Zhihedao Group, the "Due-Diligence-Officer" is an operating system for due diligence interviews designed for investors. It uses evidence strength as the criterion for decision-making. It features a built-in four-step risk pricing method (calculating risk value via Probability × Impact, converting to valuation discount ranges and term protections) and seven-dimension, three-layer inquiry (Integrity, Cognition, Self-iteration, Leadership, Stress Resistance, Benefit Sharing, Motivation/Mission; weighted scoring determines grades A–E, with five red lines triggering a one-vote veto). Evidence gathering is divided into six engines: Finance, Legal, Business, Team, People, and General Interview. Pre-investment, it routes through 11 due diligence archetypes × 33 hypothesis cards to check three key items first based on institutional style; post-investment, it utilizes a hypothesis ledger state machine and a pitfall map to create a calibration closed loop. Deliverables include interview outlines and "combat packs," review minutes, a six-part interview report, a ten-chapter risk control opinion, and investment decision memos. Each conclusion is categorized as Verified, Pending, or Doubtful with source citations; red lines are distinguished as absolute or relative; and conditional clauses are written with three elements: materials, time limits, and acceptance criteria. It is suitable for investment partners, investment directors, due diligence heads, investment managers, analysts, buy-side FAs, and lawyers. It does not issue investment proposals, replace conclusions from licensed professionals, or design non-compliant transaction structures; valuation methodology, term design, IPO paths, and BP pre-screening are referred to corresponding skills.
Read More墨客排版: Delegate WeChat Article Formatting to Skill
Inker Layout (inker-layout) is a WeChat Official Account content layout Skill published by user_6efee5b0 on the SkillHub community. Currently at version 1.1.1 and categorized under content-creation, it requires no API key and has achieved a page rating of 100,000 with 896 downloads. It covers graphic and text layout, visual design, and SVG interactive components, featuring ten built-in style themes. It utilizes a three-stage Q&A guided workflow: after being triggered by trigger words, it confirms requirements in multiple rounds before generating the layout. The article details its positioning, core capabilities, universal SKILL.md-based installation process, and typical use cases. Furthermore, it highlights usage precautions, including SVG component compatibility, community sourcing, and zero installations, while providing the official URL.
Read MoreA-Share Dual-Blade Stock Selection: Setting Direction Pre-market, Validating Capital Intraday
The "Mulan" event-driven stock selection skill integrates a pre-market media matrix with real-time anomalous capital flow verification via a dual-source serial approach. It excludes directions susceptible to post-news decline (sell-on-news) and filters individual stocks through a multi-tier funnel, outputting a watchlist with pre-defined exit strategies. The system features Fast/Full modes, an emotional gatekeeper, and batch parallel processing capabilities.
Read Moredsh-task-notifier: System Notifications on Task Completion and Automatic Retry on Failure
dsh-task-notifier is a DSH Host workflow plugin that monitors turn completions and approval events. It delivers system notifications on Windows, macOS, and Linux (utilizing the DSH icon for Windows Toasts) and, based on configuration, automatically sends a "continue" command in the event of network or model request failures. Drawing on the official README, this article explains desktop patch mounting, the "dsh plugin add" installation method, and configuration details such as "autoContinue". It is compatible with DSH versions >= 0.2.1。
Read MorePaddleOCR Multilingual Document OCR and Structured Data Toolkit: Convert Images and PDFs to LLM-ready JSON and Markdown
This Skill encapsulates Baidu's PaddleOCR into a generic, SKILL.md-invocable capability designed to convert documents, images, and PDFs into structured, AI-friendly JSON or Markdown. It supports over 100 languages, facilitating integration with LLMs for extraction, retrieval, and summarization tasks. As a community-sourced project (Version 1.0.0), it is categorized under "office-efficiency" and tagged with "requires_api_key=false". Since reproductive installation commands, configuration fields, and official command-line examples are not provided in the documentation, it should be enabled via the SkillHub page or coordinates. Users must verify PaddleOCR dependencies, complex layout recognition, and actual performance within the target client before use. Be aware of community supply chain risks; do not judge quality solely based on ratings or download counts, and do not assume undeclared features such as translation or typesetting generation.
Read MoreCodex Cleaner: Secure Auditing and Purging of Codex Local Storage in the ChatGPT Desktop App
Codex Cleaner is a cross-platform skill for securely auditing and cleaning Codex's local storage in the ChatGPT desktop app, covering Windows and macOS, with a focus on the .codex directory, Codex runtime/cache/temp remnants, old daily report thread archiving, and retention policies. It supports checking disk usage, cleaning low-risk temporary files, archiving old daily reports, deleting expired daily report sessions, and designing secure retention policies. Maintained by community publisher user_433daef8, version 0.1.13, categorized as it-ops-security with the tag requires_api_key=false, distributed via SkillHub at @user_433daef8/codex-cleaner-romancekami. This article covers its positioning, core features, installation and activation methods, typical usage, applicable scenarios, and considerations.
Read Moredsh-root-temp-guard: Turning the "temp files in tmp" convention into runtime interception
dsh-root-temp-guard is a DSH Host workflow plugin maintained by lizhijie429. It intercepts or quarantines temporary directories and files under the workspace root at three layers: before tool execution, on fs write/edit intent, and at the end of a round. It supports configuration options such as tmpDirName, mode, and quarantine. It serves as a convention and hygiene guard rather than a security boundary. Recommended CLI installation: `dsh plugin --profile web add dsh-root-temp-guard`. Suitable for agent development scenarios that already maintain a standardized `tmp/` directory and want to reduce root-level pollution.
Read MoreMastering Lyric Writing: Transforming Fragmented Thoughts into Complete Lyrics
"Master Lyric Writing" is a lyric-creation Skill released by community developer user_14a81405, version 2.0.1, in the content-creation category, with no API key required. Based on the complete 19-chapter methodology in Noema Lab's "Lyric Writing Teaching.md", it divides lyric creation into six stages: inspiration mining, parameter setting, structure design, first-draft generation, layered polishing, and final-draft analysis. It enables the model to actively guide users from fragmentary ideas to a fully finalized set of lyrics, rather than outputting the finished product all at once. Official Chinese trigger words are provided: write lyrics, create a song, write lyrics, fill in lyrics, help me write a song, write a song about XX, etc. The article explains its positioning, the six-stage workflow, key points for coordinates and configuration, invocation examples for idea-level input, as well as applicable scenarios and limitations that need to be checked on the page before use.
Read Moredsh-push-privacy-guard: Privacy check before DSH commit and push
dsh-push-privacy-guard is a DSH Host plugin that performs read-only scans of the staging area and push scope before an Agent executes git commit, git push, or gh repo sync. It detects sensitive information including API keys, private keys, connection strings, and local paths. When a match is found that is not on the whitelist, it requests user authorization through the DSH approval panel; if approved, the command is allowed and written into a manageable masked whitelist; if rejected, the command is blocked. It supports optional full-repository scanning on first push, the /privacy-guard slash command, and hot-update configuration. It is positioned as a content-leak guardrail rather than a complete security boundary. Installation is available via the dsh plugin add GitHub source or a custom marketplace index URL.
Read MoreBilibili API Service: MCP Server for Integrating Bilibili Content Capabilities into Agents
"Bilibili API Service" is an MCP server for the Bilibili API, identified by the coordinates @org-rn88lg3j/bilibili-srv. Version 1.0.0, published by Xiao Ben Yang Technology as an enterprise source and categorized under data-analysis, it encapsulates common Bilibili data operations—such as video search and user content retrieval—into MCP tools for Agent invocation. It is suitable for Bilibili content management and data analysis scenarios. This Skill is tagged with `requires_api_key=true`, requiring a Bilibili API key for use. Public records do not provide directly verifiable installation commands, configuration fields, or reproducible invocation examples. This description is based solely on the official introduction regarding its capabilities and does not involve unmentioned features such as comments, bullet screens (danmaku), or live streaming. It is recommended to verify the official page and evaluate stability and data compliance before integration. The official address is https://api.skillhub.cn/org-rn88lg3j/bilibili-srv, and the SkillHub display page is https://www.skillhub.cn/skills/org-rn88lg3j/bilibili-srv.
Read More