Word Batch Formatting Unification Tool: Unify Formatting for Large Volumes of Documents with VBA Macros
This article introduces the Community Skill Word Batch Formatting Tool (@user_9355e276/word-format-batch-adjust, version 1.0.3, categorized under office-efficiency). It utilizes VBA macros to batch-unify fonts, paragraphs, margins, and heading styles across multiple Word documents. This tool is ideal for the standardization of large volumes of documents such as reports, bid documents, and contracts, serving as an efficient alternative to manual, document-by-document adjustments. The text details its functional scope, ratings and download statistics on SkillHub, verification constraints regarding installation and activation methods, and its operational workflow. It also highlights essential precautions, including the requirement for local Word installation and macro permissions, the necessity of backing up files prior to execution, and dependencies on existing heading styles, while providing the official access URL.
Read Moredsh-code-intel: Read-only Semantic Navigation with Local LSP
dsh-code-intel is a DSH client plugin maintained by lengrufeng2008. It provides code_definition, code_references, code_symbols, code_diagnostics, code_hover, and preview-only code_rename through a local Language Server, with no embedding or remote indexing. Installation requires `dsh plugin --profile web add` along with the corresponding language LSP. Paths are restricted to the workspace by default, and positions are 1-based. MIT licensed, requires Node >= 20.
Read Moredsh-action: Cost-Capped DSH PR Review in GitHub Actions
`lengrufeng2008/dsh-action` is a DSH community workflow plugin that, when triggered by `pull_request`, uses headless DSH with the `pr-review` skill to review the diff, generates Markdown, and can post a comment on the PR. The article covers Action configuration, `DEEPSEEK_API_KEY`, the `max-minutes` wall-clock limit, and security considerations, and provides a local bundle installation example. MIT licensed; it is not a merge or security audit tool.
Read MoreIn-depth Government and Enterprise Visit Strategist: Preparation Skills for Account Managers
Introducing a Skill created by community developer @user_9eb1724b: “Government & Enterprise Deep On-site Visit Advisor” (政企深度走访智囊). Version: 1.0.3. Category: business-ops. Location: @user_9eb1724b/telecom-visit. Tag: requires_api_key=false, no additional API key needed. Verified capabilities include customer industry pain-point analysis; tiered visit scripts (initial icebreaker / needs deep-dive / business negotiation); explanations of the value of mobile digital products; follow-up and revisit plans; and marketing compliance reminders. Trigger keywords: visit preparation, on-site visit, client negotiation, enterprise research, visit scripts, visit plan, negotiation preparation, visit outline. Not applicable to: internal report/statistics, full-text proposal writing, or discount approval process inquiries. The official documentation provides no command-line installation command or differences in integration for Cursor/Codex CLI/Claude Code; the body text describes this as-is and does not fabricate cases, data, or user feedback.
Read MoreEconomics · Mentor Gao Wei: Comprehensive Logic-Based Test Prep Coaching Skills
"经济学·高维导师" (Econ Master Mentor) is a community Skill on SkillHub (@user_fa14b5bc/econ-master-mentor, v1.0.0, category: Education, tag: requires_api_key=false), covering seven major knowledge domains: microeconomics, macroeconomics, public finance and taxation, monetary and financial studies, statistics, accounting, and law. Its core capabilities include extracting key points, building conceptual frameworks, thoroughly explaining underlying logic and economic intuition, analyzing common exam pitfalls, and diagnosing learning obstacles. Following a "explain thoroughly first, then address exams" philosophy, it does not speculate on exam questions and does not replace textbooks or syllabi. The article outlines its positioning, trigger word design, installation source, and typical queries routed by the invoking Agent. It also emphasizes that chapter details and installation steps should refer to the official documentation.
Read Moredsh-runner: Centralized Management of Development Services' Run Panel in DSH
dsh-runner is a DSH client plugin maintained by lcx19970714 (v0.17.1, MIT). It lets you start and stop development commands grouped by workspace in the left-hand "Run" panel, and supports dependsOn, a strict ports contract, logs, and port diagnostics. It also provides `dsh_runner_config` / `dsh_runner_run` so AI can maintain configuration and manage processes. Compatible with the DSH 0.2.1 line; Windows + PowerShell only. To install, paste `github:lcx19970714/dsh-runner` on the plugin page, or run `dsh plugin --profile desktop add`.
Read MoreClassroom Observation and Evaluation Skill: Structuring Observation Records and Evaluation Dimensions
Class Observation & Evaluation (@kunlungrowth/edu-lesson-observe, v1.0.0) is an enterprise-source Skill published by Kunlun Growth under the SkillHub education category, positioned as a framework for teacher class observation records and dimension-based evaluation analysis. It offers two core capabilities: first, a structured observation framework that organizes scattered observations—such as blackboard writing sequences and questioning frequency—into specific teaching segments, teacher activities, student activities, and time allocation, preventing them from becoming unstructured free-flowing notes; second, a checkable evaluation checklist that allows users to categorize facts by dimension and flag doubtful points after class. Based on the standard SKILL.md format, this Skill is tagged with requires_api_key=false and does not rely on external API keys. The SkillHub page displays a rating of 100000, 981 downloads, and 0 installations. The text notes that the official page was inaccessible during this review; thus, there are no verifiable installation commands or configuration items, though it was confirmed that no API key is required and it can be loaded via the Skill directory. The typical usage section represents a general workflow inference based on its positioning rather than an official example. Users should visit the official page to verify framework items and the version prior to use.
Read MoreError Journal Skill: Distilling Questions, Follow-ups, and Root Causes into Searchable Local Archives
Mistakebook (mistakebook @user_dad39c04, v1.0.0) is a community education Skill on SkillHub that requires no API key. It analyzes uploaded question images or screenshots to reconstruct and explain options item-by-item, address learner inquiries, fully retain multi-turn Q&A context between the user and the AI, locate corresponding book sections, diagnose error causes, and write structured metadata along with original image bytes into a local searchable archive. It is suitable for learning agents requiring long-term accumulation of mistakes, preservation of Q&A context, and mistake review. Since official documentation lacks command-line installation steps and reproducible examples, implementation requires verification in specific environments prior to integration.
Read MoreISO Standards System Certification Assistant: A Companion Skill for Multi-Standard System Certification
The **ISO Standard System Certification Assistant** is a "Professional" category Skill published by SkillHub community user @user_7ad16bf9 (Version 1.0.0, Source: Community). It does not require an API key (requires_api_key=false) and is currently rated 100,000 with 937 downloads and 0 installations. Designed for the entire workflow of ISO system certification (covering standards such as 9001, 14001, 45001, 27001, 20000, 22000, 50001, 13485, 22301, 27701, 37301, 17025, 37001, 55001, and 21001), it integrates clause interpretation, analysis of existing documents, gap analysis, generation of standardized procedural documents, automatic conversion of legacy documents to audit-ready formats, and simulated audit Q&A into a cohesive pipeline. The article follows a fixed structure to introduce the topic: "What is it," Core Features, Installation & Activation (noting that the official page and API data fetch failed during this capture, so installation commands and configuration fields remain unverified—stated factually), Typical Usage (with examples from the official introduction), Applicable Scenarios & Considerations, and a Brief Conclusion with Official Links.
Read MoreComprehensive Resume Analyzer: A Deep Health Check for Resumes Without a JD
This article introduces the SkillHub community skill 'Resume All-in-One Analyzer'. It does not require uploading a job description (JD). Once a resume is provided, it performs four key tasks: (1) conducts a three-tier diagnosis (critical/important/minor) of issues and provides side-by-side comparison from original wording to revised phrasing; (2) assesses career path continuity and identifies growth stagnation risks; (3) constructs a growth tree and recommends Top 5 matched positions; (4) outputs an ATS compatibility check, HR background check rehearsal, and a job application progress tracking template. Coordinates: @user_068c5cf6/resume-analyzer-v1. Source: community. Category: content-creation. Version: 1.1.2. No API Key required. This skill is ideal for resume optimization for job hunting, career direction uncertainty, interview preparation, and cross-industry transition assessment; while for JD-targeted resume optimization, please switch to resume-optimizer.
Read MoreDSH LeRobot Dataset Inspector: Read-only Viewing of Local LeRobot Datasets in Harness
dsh-lerobot-dataset-inspector, a DSH client plugin maintained by Lavalx100, offers the lerobot_inspect_dataset profile command that provides read-only summaries of LeRobot v2.0, v2.1 and v3.0 datasets via the active file system, including version, features, tasks, episode and asset counts. An optional episode_index argument displays single-episode metadata and per-episode statistics for v3.0. It does not open robot Parquet files or decode video streams. Requires Node ≥24.9 and DSH 0.2.2-alpha.2 up to (but not including) 0.3.0. Install by adding a local bundle path with the dsh plugin --profile command. Scan and vectorisation outputs are bounded; when the limit is reached, the results are flagged as incomplete.
Read Moredsh-skill-prompt-injector: Search, pin, combine, and inject Skills and Prompts in the DSH right sidebar
dsh-skill-prompt-injector (Larliers, v0.2.0, MIT) is a DSH client plugin that mirrors the current session's skill directory in the right sidebar, managing prompts and combinations. It supports global pinning and one-click appending to the composer. It requires DSH >=0.2.1-rc.3 <0.2.2 and the `sidebar.right.pane.tab`. Recommended installation: `dsh plugin --profile <profile> add dsh-skill-prompt-injector`. After restarting, use the 'Skills & prompts' tab. Data is stored by default in `~/.dsh/skill-prompt-injector/`.
Read MoreSummarizing Webpages, Podcasts & Videos | Jianshi AI: Transforming Long Content into Actionable Summaries & Checkpoints
Introducing the Skill "Web, Podcast, and Video Content Summarization | JianShi AI" (located at @user_fccf428b/china-content-summarize, version 1.0.0, categorized under knowledge-management, community-sourced). It targets enterprise content and new media teams, transforming web pages, podcasts, and video materials into three types of structured output—actionable plans, checkpoints, and risk boundaries—for use in official accounts, official websites, knowledge bases, sales letters, and brand content production, without requiring API Key configuration. The article explains the core features, official address, typical usage, and applicable scenarios, and notes limitations such as the official documentation not providing installation commands, zero installations, and reliance on transcription for input quality, recommending that users first open the official address to confirm the current instructions before use.
Read MoreOne-Page Procurement Requirements Summary
# Introduction While submitting a procurement request may seem simple, the actual execution often leads to disorganization. During collaboration between procurement and R&D, the requester often casually mentions "Need to purchase XX, approximately XX units" in a chat message. However, specifications are incomplete, quantity metrics are inconsistent, the usage is not explained, and the budget scope is unclear. Consequently, collaborators in procurement or finance are forced to engage in back-and-forth follow-ups. Ironically, the cost of this verification exceeds the effort required to write a formal document in the first place. This type of information needs to be consolidated into a standardized, structured specification, but establishing a dedicated process for it is too cumbersome. The "Procurement Request One-Pager" Skill introduced below is designed to solve exactly this: it consolidates scattered procurement information into a single-page specification that clearly outlines specifications, quantities, usage, and budget parameters. # What is this? The "Procurement Request One-Pager" is a community-sourced general Skill, located at `@user_bd3231bd/procur...`
Read MoreSpoken Script Generation: Entrust Short Video Sales Scripts to an Agent
Short-form Video Voiceover Script Generation is a content-creation Skill published by Kunlun Growth on SkillHub. It is located at @kunlungrowth/kouchong-script, version 1.0.0, released under an enterprise source, and does not require an additional API key. What it does is generate natural, fluent short-video voiceover scripts and on-camera delivery, removing boilerplate while preserving a human touch; it is suitable for general product-pitching scenes. The article outlines its positioning and the verified capability line, notes that the official page’s body text is currently not retrievable, so the install command, directory layout, and configuration fields could not be verified. It only provides identifying information such as location, version, source, and category, and reminds readers that the metrics—930 downloads versus 0 installs and a rating of 100,000—should not be used as selection criteria. The claim of being “human-like” lacks quantifiable criteria, and manual spot-checks are recommended. It is suitable for trial runs by e-commerce creators, MCN content teams, and developers who need to batch-produce voiceover scripts within agent workflows.
Read Moredsh-ledger: DSH Local Personal Bookkeeping and Spending Analysis
dsh-ledger is a personal accounting plugin developed by laomaogu for DeepSeek Harness. All data is stored locally at ~/.dsh/personal-ledger/ledger.json. The dashboard provides three pages — Overview, Details, and Settings — featuring KPIs, five types of inline SVG charts, and up to six budget/spending heuristic tips. A unified Host-level buildSnapshot ensures consistency with the Agent tools ledger_add, ledger_report, and ledger_delete. It can be installed via the Web plugin manager or `dsh plugin add` using a local directory or github:laomaogu/dsh-ledger. Licensed under MIT 1.0.0, it requires Node ≥22.19 and dsh ^0.2.1-alpha.2.
Read MorePatent Abnormal Risk Assessment: Screening Abnormal Applications Item by Item in Accordance with CNIPA Order No. 77
Patent Non-Normal Risk Assessment (@user_07bb1a57/patent-abnormal-assessment, v1.0.0, community/professional) scrutinizes one or a batch of patent applications item-by-item against the constituent features of non-normal application behaviors stipulated in Article 3 of the *Provisions on Regulating Patent Application Behaviors* (Order No. 77 of the CNIPA). It outputs risk signal intensity, evidentiary completeness, a checklist of materials to be supplemented, and appeal pathways following the issuance of a notification. This tool is applicable for pre-application self-examination, agency compliance self-audits, client risk profiling, and responses to notifications regarding non-normal applications. It is not applicable for patent valuation, patent rights stability assessment, novelty search, or the drafting and proofreading of application documents.
Read Moredsh-remotion-launcher: One-click Management of Remotion Studio Animation Projects in DSH
`dsh-remotion-launcher` is a DSH workflow plugin maintained by laojingwei. It does not include the animation project itself but uses a floating panel in the Web interface to manage the installation path, Vite builds, TSX server start/stop, and health checks for `remotion-studio-mcp` projects. It also enables one-click copying of stdio or HTTP MCP connector JSON to DSH. The plugin and the project are two separate packages requiring DSH >= 0.2.1-rc.2, with configurations stored in `~/.dsh/remotion-launcher.json`. Based on the official README, this article introduces installation commands, typical usage, and environment variables, while also explaining permissions and known mutually exclusive behaviors.
Read MoreFeishu Mail Assistant: Delegate Feishu Mail Daily Operations to a Skill
Feishu Mail Assistant (@user_af28adda/feishu-mail-assistant) is an intelligent Feishu mail assistant Skill built on lark-cli, published and maintained by community developer user_af28adda. It is categorized under office-efficiency with version 1.0.0. It encapsulates repetitive email operations—such as email triage, summarization, follow-up tracking, mail merge, attachment archiving, and email-to-task conversion—into structured workflows. The Skill orchestrates the upper-layer workflows, while the underlying email operations remain executed by lark-cli, thereby avoiding scattered implementations in ad-hoc scripts. The Skill defines Chinese and English trigger words (e.g., 邮件分拣, 邮件摘要, 未回复追踪, 批量发邮件, 附件下载归档, 邮件转任务, as well as mail assistant, email triage, follow-up, mail merge, etc.) and includes a metadata tag of requires_api_key=false. This article outlines its positioning, capabilities, activation methods, typical use cases, applicable limitations, and provides the official link.
Read MoreDocument Analysis MCP Server: 21 MCP Tools for Document Statistics and Summarization
Released by user_053d27d0 (Version 1.0.1), the Document Analysis MCP Server originates from the 'community' source under the 'knowledge-management' category. It encapsulates eight common document analysis tasks—file reading, word frequency counting, TF-IDF keywords, TextRank summarization, sentiment analysis, readability scoring, statistical analysis, and anomaly detection—into 21 MCP tools that Large Language Model (LLM) clients can invoke on demand. This skill requires no external API keys, though its installation count remains at zero. This article has verified the capability list provided in the official description; however, the installation command was not provided due to a failure in scraping the official page. It is recommended to verify the tool names and configuration methods on the official page.
Read MoreIndex Express ETF Information Query: On-exchange ETF Basic Information and Market Data Query Skill
Introducing the Skill "Index Express ETF Information Query" released by E Fund Management Co., Ltd. (ETB Fund Management Co., Ltd., version 1.0.3, source: enterprise, category: professional, requires API Key configuration). This Skill allows querying of on-exchange ETF basic information, fund size and fees, holding details, historical dividends, period returns, fund flow, top gainers, hot search rankings, reverse ETF search by held stocks, as well as latest price, intraday change, trading volume, turnover rate, IOPV, premium/discount rate, and minute-level trends by 6-digit security code (e.g., 510300). It is suitable for scenarios involving on-exchange ETF basic information, trading activity, theme screening, peer comparison, and real-time market data; for off-exchange subscription and redemption of index funds, use open-fund-query instead; for querying the index itself, use index-query instead. The official page failed to be retrieved, so installation commands and configuration fields could not be verified; this has been honestly noted in the main text.
Read MoreDetailed Judgment Analysis Skill: Automation from Legal Documents to Client Reports
Judgment Analysis Skill is a professional-category Skill published on the SkillHub community. Located at @user_6e8ac8d8/ecj, it is version 1.0.0 and does not require an API Key. It processes judgments, rulings, and mediation documents from Chinese courts, with inputs supporting PDF, DOCX, scanned images, and text. The workflow begins with a preliminary client interview to confirm the client, service objectives, procedural status, and the intended recipients of deliverables. It then extracts claims, the focal points of dispute, the court’s reasoning, the operative part of the judgment, monetary amounts, time limits, and follow-up actions. Finally, it produces attorney-verified materials with a business-style cover, as well as visualized HTML/PDF reports tailored for general clients. The skill focuses on common civil cases; for other case types, it applies a safe fallback protocol. It is designed for use cases such as determining case types, explaining judgment outcomes, preparing client reports, case-closing reports, appeal evaluation materials, reminders for performance of obligations, and enforcement-preparation checklists.
Read MoreWorkflow for Development of Safety Training Materials: Full-Process Automation for Training Materials for Gas Stations and LNG Stations
A Skill for producing training materials in the domain of safety production for gas stations and LNG refueling stations, published by Safety Engineer Chen, version 1.4.1, category Education, source Enterprise, no API key required. Its core capabilities include training needs analysis, PPT generation, handout generation, and quiz generation, with review for professional accuracy to enable end-to-end automation from requirements to delivery. The article describes its positioning, capability boundaries, how to enable it, a typical usage workflow, and notes on domain limitations and the need for manual review before delivery.
Read MoreComprehensive Insurance IP Strategy | Client Acquisition & Recruitment: An Offline-Ready Toolkit for Personal Branding Tactics
An all-process playbook for insurance IPs targeting agents, brokers, and small team leaders, featuring dual strategies for client acquisition and team recruitment. Comprising nine chapters ranging from positioning to review, it includes six bottleneck diagnostics, a terminology dictionary, 14 scripts for handling objections, a 100-day action plan, and three livestreaming compliance gates. It comes with a local checking script `insur_check.py` (685 lines) that operates offline without calling external models, offering five subcommands: --template, --diagnose, --topics, --objection, and --audit. The --audit command screens content for income promises, medical claims, and false urgency, while verifying the presence of next steps and client-centric language. It provides line-by-line error prompts for incorrect parameters. A dedicated chapter outlines compliance red lines: no return guarantees, no specific security recommendations, no crossing medical boundaries, and no pyramid-style recruitment. This is designed for practitioners committed to building an insurance IP through proven methods, not for those seeking overnight viral success.
Read Moredsh-whitelist-auto-approval: Append writable roots by command to reduce danger-full-access prompts
This article introduces the DSH plugin, *dsh-whitelist-auto-approval*, maintained by laoer. It retains the workspace within the kernel sandbox for whitelisted CLIs (such as `lark-cli` and `gh`) and appends configurable writable roots, thereby avoiding the need for repeated requests for `danger-full-access`. It explains the differences between this plugin and those that allow external execution or grant global writable directories, outlines the strict matching rules, describes the UI and npm installation methods, and discusses the narrowing of roots and its applicable boundaries.
Read More