Word Batch Processing · Formatting and Layout Skill: Batch Modify Formatting, Replace Content, Extract Information, and Format Official Document Layout
Word Batch Processing · Formatting & Layout Skill, released by Changsha Xingwang Intelligent Technology Co., Ltd., is currently at version 1.2.1. Categorized under office-efficiency, it does not require an API Key. It provides 8 local batch processing functions: batch find and replace across runs, unified font/size/spacing, extraction of text and tables to CSV, batch insertion of headers and page numbers, splitting by headings, merging multiple documents, batch document generation from templates with lists, and Markdown to Word conversion. All processing is performed locally without relying on the Word client. Note: Word to PDF conversion and automatic table of contents generation are not supported.
Read MoreInterpretation of Learning Reports for Parents: Transforming Academic Performance Data into Accessible Interpretations and Actionable Recommendations
Learning Report Interpretation (Parent-Oriented) is an education-category Skill published by Kunlun Growth on SkillHub, with coordinates @kunlungrowth/kunlun-edu-report-interpretation and version 1.0.0. Targeted at parents and homeroom teachers, it transforms grade reports, learning analytics, exam data, and class quality assessments into easy-to-understand, anxiety-free interpretations and actionable home-learning recommendations. It is applicable to parent-teacher conferences, school-home communication, and transitions between academic advancement and family education. The official description marks requires_api_key=false, indicating that no additional API Key is required. Among the retrieved materials, no official installation commands, configuration fields, or reproducible usage examples have been provided; this article reports this fact as-is without making any assumptions. Before use, it is recommended to refer to the official page for the most accurate information.
Read Moredsh-ui-probe: Automated acceptance of graphical plugins in the real DSH interface
dsh-ui-probe is a DSH workflow plugin maintained by yuexizhishang. It opens target plugin pages within the currently running client to read layouts and computed styles, perform clicks and form inputs, wait for asynchronous results, and capture console errors, automatically returning to the original conversation upon completion. It can be installed via `github:yuexizhishang/dsh-ui-probe` or `@yuexizhishang/dsh-ui-probe` without requiring additional configuration. It provides the `ui_probe` and `ui_page` tools along with documentation on security boundaries. It is suitable for graphical plugin developers to perform acceptance testing; please review the source code and MIT license before installation.
Read Moredsh-columbina-pet: Columbina 2D Desktop Pet on DSH
dsh-columbina-pet is a DSH plugin maintained by Yuanawaqwq that turns Columbina into four 2D Live desktop pets. It features 76 original lines in Chinese, Japanese, and English, scene-aware chatter, weather broadcasts, realistic physics, and asset self-check. Derived from a 2D branch of dsh-pet, it is MIT-licensed and currently at 1.2.1. The article outlines core capabilities, how to add github:Yuanawaqwq/dsh-columbina-pet via DSH or install the profile via npm, typical right-click usage, and healthz checks for troubleshooting, and reminds users to fully restart DSH, mind process permissions, and respect asset licensing.
Read MoreFull Process of Medical Damage Liability Disputes (by Lawyer Zhang Chenlu): Integrating Case Handling Procedures into a Skill
This SkillHub community skill, “Medical Damage Liability Disputes – Full-Process Workflow (by Attorney Zhang Chenlu)”, is located at @user_281f816d/medical-dispute, version 1.0.1, and categorized as professional. It breaks the handling of medical damage liability dispute cases into executable steps: medical record review plus a dedicated e-medical-record module → appraisal/hearing → fault analysis → damages calculation, and it also covers medical aesthetic disputes, expert assistants (party-appointed expert witnesses), mediation by the Medical Disputes Mediation Committee, and a quick-reference for nationwide compensation standards. The tag requires_api_key=false indicates no API key is needed to enable it. Installation commands and configuration fields are not provided in the materials and should be confirmed on the SkillHub page. As a community-sourced skill, its rating and download counts are point-in-time data captured at crawling.
Read MoreBinance Futures Trading Assistant: USDT-M Perpetual Analysis and Execution Assistant Integrated with Local trading_engine
Introducing a Binance USDT-Margined perpetual contract trading assistant that integrates with the local trading_engine: it generates leverage and USDT position sizing plans based on account balance and holdings, and trade opening and closing are executed automatically by the engine, with push notifications issued accordingly. Maintained by community developer @user_ce40e76c, version 1.0.0, source community, category professional, tag requires_api_key=false, only for contract trading, not spot trading. It covers analysis, trade opening, trade closing, copy trading, order placement, take-profit and stop-loss levels, BTC/ETH trading plans, and other trigger scenarios, and provides reproducible question examples. Due to inaccessibility of the official page, installation commands and configuration fields have not been verified, and this is explicitly stated in the text for accuracy.
Read MoreDSH Message Branch: In-place edit query and branch resend
dsh-message-branch is a DSH workflow class preview plugin maintained by yoinchen23-oss. It provides a pencil icon entry next to the user message copy button, allowing in-place modification of plain-text prompts. After the original session stops and the queue clears, it creates a new branch via `sessions.fork` and sends the message, leaving the original session and history unmodified. It only supports plain text; messages with attachments are prohibited from saving. Install via `dsh plugin add` using a local repository or the tgz path from a Release. Verified on DSH 0.2.1-rc.5 Desktop. MIT License, no telemetry.
Read MoreA-Share Sentiment Cycle Assessment: Locating Phases via Intraday Snapshots
This Skill was published by @user_83e3f538 on SkillHub, sourced from community, categorized as professional, and is at version 1.2.4. It transforms the question "How is today's market sentiment?" from a subjective judgment into reproducible, phased conclusions by performing four steps in sequence: sentiment cycle positioning (freezing point, repair, fermentation, climax, divergence, and ebb phases, based on maximum consecutive limit-up height, failed limit-up rate, and large loss/large profit effects), deep diagnosis of overall market sentiment (index levels and trading volume, advancing/declining stock counts, limit-up/limit-down structure, northbound capital flows), sector strength ranking (price gains, number of limit-up stocks, net capital inflow), and leading stock identification (market position, recognition, price action characteristics). Metadata shows requires_api_key=false and pricing_type=alipay; no API Key is required. The official page does not provide command-line installation commands or manual configuration options; refer to the current page display for details. It performs only objective data analysis, does not recommend stocks, and does not constitute investment advice; the reliability of conclusions depends on the completeness of the input snapshot, and if items are missing, the corresponding sections will lack supporting evidence and must be completed before execution.
Read Moredsh-refkeeper: Validate and Export Cited References from DSH Conversations
dsh-refkeeper is a networked plugin for DeepSeek Harness, maintained by ymwang92. It parses DOIs, arXiv identifiers, ISBNs, and APA/GB/T 7714/IEEE citation lines from selected conversation ranges, validates metadata against Crossref, OpenAlex, arXiv, and DataCite, exports validated entries as BibTeX/RIS with optional generation of abstracts and conversational source sentence evidence, and moves unconfirmed entries to quarantine. Add dsh-refkeeper in DSH Settings → Plugins to enable it. Output is located in the refs/ directory of the conversation workspace.
Read MorePerformance System Compliance Reviewer: A Skill for Clause-by-Clause Compliance Health Checks
Performance systems are a hotspot for labor disputes; if clauses regarding PIPs, elimination of the lowest performers, probation assessments, or post transfers and salary reductions are illegal, enterprises face consequences beyond just compensation. Published by community publisher user_fa14b5bc (Version 1.0.0, Category: Professional), this Skill breaks down performance system texts for clause-by-clause review. It categorizes clauses into four types—PIPs, elimination of the lowest performers, probation assessments, and post transfers/salary reductions—and provides four-level risk ratings (🔴/🟡/🟢/⚪), specific revision suggestions, rewriting examples, and specialized checklists for four major scenarios. The installation command is loaded via `/plugin marketplace add` and requires no API Key. It is suitable for HR, legal, and consulting professionals to conduct internal compliance audits. It provides compliance reviews and rewriting suggestions but does not constitute legal advice; implementation should be reviewed by a practicing labor law attorney before execution.
Read MorePolymer Materials Expert: Selection, Design, and Failure Analysis of Polymers and Fiber-Reinforced Composites
The "Polymer Material Expert" (@user_2c769e18/ploymer-material-expert, v1.0.0, community, professional) is a Skill based on the general SKILL.md format. Its knowledge base is derived from 22 authoritative documents located in the "02-Product Design/Material Science" folder of the THINKER-CONSULT-LIB. It covers the complete business chain, including material selection, formulation and modification design, structural design (CLT/failure criteria), molding processes (full spectrum of injection molding/compression molding/RTM/winding/pultrusion/thermoset & thermoplastic), performance evaluation and testing standards, failure analysis (defect/damage detection, fracture interpretation, CAI), and recycling. It supports scenarios such as polymer selection, engineering plastic applications, glass fiber reinforcement/filler modification, laminate design and strength verification, injection molding defect countermeasures, composite failure analysis, battery pack non-metallic components, automotive lightweighting solutions, cable materials, and coatings. This article outlines its capabilities and usage boundaries based on the structure of "What is this — Knowledge Base — Core Capabilities — Trigger Scenarios — Installation & Activation — Typical Usage — Applicability & Boundaries." Furthermore, it clarifies that it is not applicable to metal casting heat treatment, welding, mold flow CAE, pure structural FEA, or automotive steel selection.
Read MoreSmart Transportation Knowledge Base – National and Industry Standards: Querying GB/JT/GA Standards by Scenario when Drafting Solutions
This is a standard query Skill designed for writing Chinese smart transportation solutions, maintained by community publisher user_7c5a6bbe, categorized under knowledge-management (Version 1.0.0). Its core capability is to query referenced national and industry standards based on scenario keywords, covering scenarios such as smart intersections, intelligent parking, V2X, MaaS, ETC, smart public transit, smart expressways, smart service areas, smart traffic management, smart transportation, Vehicle-Road-Cloud integration, intelligent connected vehicles, low-altitude economy, smart logistics, and autonomous driving. It supports the GB/JT/GA standard systems. Invocation requires an API key available from solvexpert.net, classifying it as an online service with external dependencies. Since official installation commands and interface examples are not provided in the source material, this article solely outlines the prerequisites for enablement and the approach to invocation. It avoids fabricating configuration items or return samples, and reminds users that the returned checklists require verification against the current status and version years of the standards.
Read Moredsh-local-market: DSH Chinese Plugin Market
dsh-local-market is a DSH client plugin maintained by YML0823 that fetches community catalogs from awesome-dsh-plugin.com. It features Chinese language search, filtering across 23 categories, and screenshot previews, enabling one-click installation via the local dsh CLI. It can be utilized as a "Chinese Market" tab within DSH or run standalone on 127.0.0.1:7788 using Node.js 20+. To install, run `dsh plugin --profile <profile> add github:YML0823/dsh-local-market`. This is an independent community project and not an official store; please review the third-party source code and permissions before installing.
Read MoreAI TCM Wellness Consultant: Packaging Constitution Identification and Dietary Advice into a Callable Skill
Introducing the community Skill "AI TCM Wellness Consultant," which provides constitution identification, dietary advice, and wellness plans based on Traditional Chinese Medicine (TCM) theory. This is Version 1.0.0, sourced from the community, categorized under life-service, and located at @user_89b96750/tcm-consultant. It requires no API Key configuration (requires_api_key=false). Due to the official page being inaccessible, this article relies solely on pre-scraped data and does not provide unverified installation commands, configuration fields, or invocation examples. It also notes discrepancies in ratings and download counts within the source material, suggesting that users should retrieve real-time data for verification.
Read Moredsh-pet-bedroom: A Web Bedroom That Can Be Opened Independently for Desktop Pets
dsh-pet-bedroom is a DSH casual dress-up plugin that provides a web-based bedroom with an independent /dsh-pet-bedroom/ route for dsh-pet. The desktop pet moves around the room image and reuses the dsh-pet animation pool. On the desk, it can read chat service diaries and memory bank notes. It optionally supports integration with QQ (OneBot/SnowLuma) and Douyin HTTP backends. Without modifying the dsh-pet source code, it is released under the MIT license and can be installed via a one-line command using the tgz archive from Releases. It requires Node.js 20+; dsh-pet must be installed first, and the bundle order must place dsh-pet before it.
Read MoreInsurer's Little Green Book Card Creator: The WeChat Graphic Content Production Pipeline for Insurance Brokers
This article introduces the "Insurance Lite Content Creator" on SkillHub (@user_b240fe8a/xls-content, version 1.0.0, source: community, category: content-creation). It is a content-production Skill designed for insurance brokers to create WeChat "image-text/image messages," transposing the "Xiaohongshu-style product seeding" methodology to the WeChat ecosystem to address the problem of insurance content that "no one wants to click, and that fails to retain users after clicking." The article outlines its positioning and basic parameters (no API key required) and its working approach, and notes that the official page could not be opened this time; installation methods and specific workflow details should be based on the SKILL.md inside the Skill. It also cautions that this Skill is a community release, shows zero installs on the page, lacks public usage feedback, and that insurance content should still be independently reviewed prior to publication in accordance with platform rules and business compliance requirements.
Read MoreWPS Lingxi Check-in for Points: Automating Daily Sign-ins with Browser Automation
This is an Agent Skill for SkillHub (@user_d550b199/lingxi260814, v1.0.2, community source, categorized as ai-agent), positioned as a daily check-in assistant for Lingxi (WPS AI). It utilizes browser automation to complete check-ins and claim Lingdian points, implementing idempotent processing to prevent duplicate claims. The skill supports both a native Lingxi browser connector and a newly added headless browser adaptation. As the official page was JavaScript-rendered and the body content could not be scraped, claims in the task description regarding the Agnes execution engine, trigger words, and zero Lingdian consumption could not be verified. Furthermore, installation commands and configuration fields (only confirmed as requires_api_key=false) were unavailable. The text only provides the official address and suggests referring to the page for details. The recorded page data shows a rating of 100000, 1235 downloads, and 0 installations.
Read More**dsh-sticker-vault: Local Collection and Contextual Retrieval of Group Chat Stickers**
The DSH memory plugin dsh-sticker-vault stores group-attached images on disk keyed by SHA-256 hashes for deduplication and counting; once a threshold is reached, images are automatically archived into the vault with their author, context, and tags recorded. A Web panel manages the vault and candidates, and agents retrieve context-appropriate Markdown image lines via sticker_pick. All data is stored locally at ~/.dsh/sticker-vault/ with no network access required. This article covers installation commands, tool usage, eviction and locking rules, and known limitations.
Read MorePublic Account Image Production Line: Application and Composition of Quotas in Civil and Municipal Engineering
The "Official Account Image Production Line" is a community Skill on SkillHub (Version 3.5.0, Publisher: user_eb2234bc, Category: Office Efficiency). Targeting civil and municipal engineering projects, it delegates quota application, bill of quantities pricing, comprehensive unit price composition, and price adjustment analysis to an Agent, producing pricing sheets and explanatory notes. It requires no API key and operates on a pay-per-use basis at ¥0.20. This article covers its functional boundaries, access methods, typical workflows, and precautions: pricing results require manual review, quota versions are subject to the project's local regulations, and given the Skill's limited public documentation and zero installation count, verification is recommended prior to use.
Read Moredsh-memory-tier: Three-tier Memory and Local Retrieval Injection
dsh-memory-tier is a DSH memory plugin maintained by yingchaoyi. It divides agent memory into three tiers: long-term, medium-term, and short-term, each with its own TTL and capacity limits. It employs local bi-tuple and tag-based retrieval, and injects memory into systemPrompt.context based on relevance and character budget. The plugin provides tools such as mem_note and mem_search, supporting locked drafts, automatic writing to the short-term tier, and mem_persona personas. It requires Node.js ≥22, DSH 0.2.x, and the web profile. Data is stored in ~/.dsh/memory-tier/store.json (MIT 0.1.4).
Read MoreWorkBuddy DingTalk Message Auto-Reply: Monitoring Unread Conversations and Auto-Processing
This article introduces a community-sourced Skill named "workbuddy中钉钉消息自动回复" (located at @user_87a16637/dingtalk-auto-replay, version 1.0.1, categorized under office-efficiency). It monitors unread DingTalk conversations: for 1-on-1 chats, it uses AI to auto-reply in the user's own identity; for group chats and specified lists, it only sends WeChat reminders without auto-replying to avoid the embarrassment of automated replies in groups. The article details the capabilities based on the dws unread interface, along with officially documented pitfalls such as SERVER__PORT conflicts, self-identification, and replying to oneself. Furthermore, it provides the official URL, SkillHub page, requires_api_key label instructions, and typical trigger usage, while noting limitations such as the requirement for an API Key and the restriction against auto-replying in group chats.
Read Moredsh-exp-lib-provider: Providing dsh-mnemon with experience library memory space and JitRL retrieval tools
The dsh-exp-lib-provider maintained by yhPrime provides an exp-lib memory space for dsh-mnemon, with data stored in exp-lib.json. Writes are performed via exp_record, while reads are executed through exp_recall or panel-based search. When paired success/failure comparisons exist, items are ranked by JitRL advantage; otherwise, ranking follows evidence strength, and  is explicitly noted as non-discriminative. Installation requires a tarball, mounting the Provider, and enabling both the plugin and the memory engine (two-layer activation). Since 0.3.0, local cluster-based routing is supported. MIT licensed; requires dsh-mnemon as a prerequisite.
Read MorePersonal Financial Planning: Integrating Budgeting, Emergency Funds, Debt, and Insurance into a Single Process
Personal Financial Planning is a life-service Skill published on SkillHub by community user user_ae636a66, with version 1.0.0, a rating of 100,000, 1,087 downloads, and no API key required. It employs five general methods—the 50/30/20 budgeting method, emergency fund planning, the debt avalanche method, the basics of index fund dollar-cost averaging, and insurance basics—to help ordinary professionals build a stress-free financial system. The article introduces its positioning, core features, installation and activation methods, typical use cases, applicable scenarios, and limitations.
Read MoreAmazon Review Scraper & Bulk Export: Batch Extract, Filter, and Export Reviews by ASIN and Marketplace
This article introduces the Skill "Amazon Review Scraper & Batch Export" (@user_507a5056/review-scraper, Version 1.5.2, Community Source, Category: Data Analysis). It enables batch collection and database storage of reviews by ASIN and marketplace. The tool supports volume control via page limits, progress monitoring, filtering by star rating and time, and CSV export. It also provides ARI Review Analysis (star ratings, keywords, trends, radar charts, VOC and deep insight reports, product change monitoring, negative review alerts and workbench, industry benchmarking, and category rankings) with Markdown/HTML report export capabilities, covering eight marketplaces: US, UK, DE, JP, CA, FR, ES, and IT. An ARI API key (ari_live_*) is required; a quote is generated before collection, and points are deducted only after confirmation.
Read Moredsh-restart-yashua: One-click Restart DeepSeek Harness
The client plugin *dsh-restart-yashua* (v1.0.16, MIT) by yashua233 provides DeepSeek Harness with three entry points: a sidebar button, the `/restart --yes` command, and the `harness_restart` command. These share a unified restart implementation and enforce confirmation before execution. Features include a pre-launch read-only health check, a three-layer fallback relaunch mechanism on Windows, and troubleshooting via `restart.log`. It declares compatibility with DSH 0.2.1-rc.5. Based on the GitHub README, this article introduces installation commands, configuration options, and typical usage scenarios.
Read More