Academic Research Assistant: OpenAlex-Based Literature Search and Review Generation
Academic Research is a skill for academic literature retrieval based on the OpenAlex API (free, no API key required), located at `@clawhub_rogersuperbuilderalpha/academic-research`, published by rogersuperbuilderalpha, version v1.0.0, MIT-0 licensed. It encapsulates the ability to retrieve papers by topic, author, and DOI, expand citation chains, deeply read open-access full texts, and output structured metadata, and provides the `literature-review.py` script to automatically perform multi-query retrieval, deduplication and ranking, topic clustering, and generate a Markdown review. This article introduces installation commands, reproducible retrieval and review workflows, and usage limitations (OpenAlex is predominantly English-language literature, full-text PDFs require manual extraction, and the review script caches files in `/tmp`).
Read MorePaper Text Abstract Generation: Decomposing Academic Abstracts into Research Questions, Methods, Findings, and Conclusions
This community Skill (@user_e8eb8000/paper-abstract-summarizer, v1.0.0, in the knowledge-management category) targets a common pain point when reading abstracts: research questions, methods, results, and conclusions are often intermingled in one paragraph, so extracting them manually requires sentence-by-sentence screening, and English abstracts need to be translated again. It combines extraction, summarization, and translation, and outputs structured results into four slots: research question, methodology, findings, and conclusions. The provided information only indicates requires_api_key=false; no installation command, configuration fields, or reproducible examples are given, and two official URLs failed to be fetched at the time of writing. Therefore, this article marks those parts as unverified and does not fabricate them. Note that it processes abstracts rather than full text; the quality of extraction and translation depends on the underlying model, and key conclusions should be verified against the original. The page shows 1,010 downloads and 0 installs, indicating a very small community usage sample, so a fallback plan should be prepared when using it.
Read MoreReference Retrieval: Cross-database Search and JCR 2026 Metric Verification
Reference Retrieval is a community Skill published by @user_9cc4d736 on SkillHub, version 1.0.0, categorized under knowledge-management. It searches for verifiable academic literature across PubMed, Google Scholar, and Web of Science by topic, sorts the results, reports Clarivate-verified JCR 2026 metrics, and delivers Word documents. It is suitable for literature evidence, database retrieval, JCR/JIF queries, impact factor and JCR quartile verification. The official page does not provide installation commands, only offers the coordinates @user_9cc4d736/reference-retrieval-skill, and does not require an API key.
Read Moredsh-cross-session-ref: Cross-session Drag-and-Drop Q&A Card with Prompt Injection
Introducing KanmanJO's DSH plugin, dsh-cross-session-ref (a memory-type plugin). It resolves the issue of reusing previously clarified Q&As and attachments across multiple sessions. By dragging cards in the panel to create references, the plugin utilizes an agent/pre-step to inject the questions, original answers, and attachments into the prompt for the next query in the target session. In the chat interface, this content appears as gray context lines rather than user bubbles. It covers installation commands (npm/GitHub/local bundle), typical usage, local links.json and colors.json files, and requires MIT 0.2.1 and DSH 0.2.x.
Read MoreShort Drama Storyboard Generator: Generating a Complete Set of Vertical Short Drama Assets from Several Parameters
This article introduces the SkillHub community skill "Short Drama Storyboard Generator" (@user_ad7d6c4e/shortdrama-storyboard-generator, v1.0.0, categorized under content-creation, requiring no API Key). By simply inputting four parameters—theme, episode duration, number of episodes, and target audience—it generates character profiles, episode outlines with pacing plans, tabular shot-by-shot storyboards, AI video image generation prompts, sound effect notes, five sets of cover titles, short video promotional copy, and a script self-check report. Designed for creators on vertical video platforms such as Hongguo Short Drama and Fanqie Short Drama, it is suitable for studios and individual screenwriters to produce content in bulk. The article provides verifiable positioning information, a list of core functions, and usage examples via triggers. It also truthfully states that existing materials do not provide installation commands or configuration options, requiring users to verify the loading mechanism according to the host Agent's SKILL.md.
Read Moredsh-chat-first: Align the Direction First, Then Let the Agent Act
dsh-chat-first (Chat First) is a DSH workflow plugin released by kankanfs. Within a single session, it defaults to a discussion mode with tools disabled, and after aligning on the goal, switches to execution mode to continue the task. The current version is 0.0.1-alpha.1, and it declares compatibility only with DSH 0.2.1-rc.5. Desktop installs the npm package dsh-chat-first@0.0.1-alpha.1 via third-party plugin installation; for the Web profile, the npm package can be installed via `dsh plugin --profile web add` with a local path. Modes are persisted in ~/.dsh/plugins/dsh-chat-first/sessions/. It is MIT licensed and performs no external telemetry.
Read MorePDF Translation · Literature Translation | Document Holmes (Layout Preserved · Bilingual Comparison)
This is a community Skill on SkillHub called "PDF Translation · Literature Translation | Doc Holmes (Layout Preservation · Bilingual Comparison)," located at @indiv-sorsor/doc-holmes, version 3.0.1, categorized under office efficiency. Its core goal is to "preserve the original layout and elements": formulas, charts, tables of contents, footnotes, and annotations are retained in the output, with each file generating one bilingual comparison PDF and one pure translation PDF. Before translating, each file undergoes a "health check" for grading—Level A (intact native text layer) proceeds with fidelity translation; Level B (noisy text layer) is translated with an attached noise report; Level C (scanned/pure image) goes through an experimental OCR channel, with outputs accompanied by a preview quality description page. The engine invokes BabelDOC (pdf2zh-next) subprocesses, utilizing your own configured OpenAI-compatible endpoints and keys; subscription endpoints with encoding are rejected by default to protect account security. The batch mode features checkpoint resumption, single-file timeouts, audit logs, and failure rollback. The article introduces the capability structure, installation and activation methods, typical commands, and batch usage, as well as limitations to note: no built-in credentials, Level C quality is for preview only, and the source is community-based with a low installation count, so it is recommended to self-test with Level A documents first.
Read MoreOutlook Email Sorting Skills: Email Organization Based on Graph and Text Analytics
The Outlook Mail Triage Skill is a community Agent Skill from SkillHub with coordinates @user_3c6cb52e/outlook-mail-triage-microsoft-graph, documented version 1.0.0, and classified under office-efficiency. It reads Outlook emails via Microsoft Graph API v1.0 and delta queries for incremental mail synchronization, applies machine learning-based priority scoring through Azure Cognitive Services Text Analytics, and then archives messages into appropriate folders based on learned patterns. Before using it, you need to prepare Graph mail read permissions and Text Analytics API call credentials—this is also indicated by the label requires_api_key=true. The scraped documentation does not include installation commands, configuration fields, or reproducible examples. When integrating, refer to the instructions inside the skill package and initially validate the scoring and archiving results within a small scope.
Read Moredsh-fatfish: Big Fat Fish desk pet in the top-right corner of the DSH input box, changing expressions based on conversation status
dsh-fatfish is a DSH client plugin maintained by kabarryo that displays an animated "big fat fish" desktop pet at the top of the conversation input card. It switches between 29 embedded WebP emojis based on nine states: standby, input, generation, awaiting approval, awaiting response, error, completion, idle, and context overflow. Installation requires DSH 0.2.x; after cloning the repository, link it in the profile's package.json and restart. There is no host-side logic or configuration. The emoji assets are licensed under CC BY-NC-SA 4.0 (non-commercial use only). This article outlines the installation steps and state determination based on the repository's README and package.json.
Read MoreCET-4 English: Self-Contained Vocabulary Recitation Skill
CET-4 English (@user_cb490415/cet4-word-memorize, v1.0.0, community/education) is a self-contained English vocabulary tutoring skill. It features a built-in recitation overview of 48 lessons from Liu Xiaoyan's *Are You Still Memorizing Words?* for CET-4, complete with pronunciations, common phrases, and new word markers, ready for out-of-the-box use. Accompanied by a local service for a reading station and vocabulary lists, it supports interactive guided memorization, the generation of overview HTML files based on arbitrary vocabulary books, and word extraction from scanned PDFs without OCR, covering the entire workflow from scanned books to vocabulary lists. This article outlines core features, installation metadata, typical usage scenarios, and limitations. While the SkillHub rating is 100,000 with 963 downloads, the installation commands and local service startup methods could not be verified due to a failure in retrieving the official page; thus, the official page should be consulted for authoritative information.
Read MoreIMA Personal Note Sync Skill: Maintaining Consistency Between IMA Personal Notes and Knowledge Base
The IMA Personal Notes Sync Skill is a community Skill published by user_6537d63b, located at @user_6537d63b/ima-notes-sync, version 1.0.0, categorized under knowledge-management. It is positioned as a bidirectional incremental sync tool between IMA personal notes and a personal knowledge base: it synchronizes all personal notes to a specified knowledge base, automatically adding missing associations and removing invalid ones when inconsistencies arise between the two sides. Trigger scenarios include "sync notes to knowledge base," "keep notes and knowledge base consistent," and "batch add all notes to knowledge base." The tag indicates requires_api_key=true, so a valid API Key is required. At the time of writing, neither the official page nor the SkillHub page was accessible; installation commands, directories, configuration fields, and client-specific differences were not included in the main text—please refer to the official page for authoritative information.
Read MoreDongfang Tianyi Medical Paper Figure Reading: Explaining Paper Figures and Charts Clearly to Medical Students
Targeting medical students with no prior background in bioinformatics or statistics, this tool interprets medical research papers and bioinformatics figures in accessible language. It begins by providing a one-sentence conclusion, followed by explanations of coordinates, groupings, and statistical information. It distinguishes between facts directly visible in the figure, potential inferences made by the authors, and content that cannot be determined solely from the figure. Additionally, it highlights common misconceptions and the limitations of the conclusions. It is applicable to Volcano Plots, Heatmaps, PCA, Enrichment Analysis Plots, ROC Curves, Survival Curves, and Forest Plots. This Skill is published by Dongfang Tianyi - Research Skill on SkillHub (Coordinate: @org-ekt13cme/medical-paper-figure-coach, Version: 1.0.1, Category: Education, Source: Enterprise) and requires no API Key. The provider, Tianyi Cloud, is positioned as a research platform for AI, Research, and Bioinformatics Clouds. The retrieved materials do not include installation commands or reproducible examples; please verify details on the official page before use.
Read Moremanualcompact-billion：Restore DSH Native /compact When Coexisting with Billion-Context
After installing billion-context, DSH's /compact and compaction backends are disabled by bili-compaction-gate. manualcompact-billion re-enables dsh-compaction-basic and dsh-command-compact at the outermost layer of the configuration waterfall, setting auto to false to retain only manual compaction, thereby avoiding conflict with billion's automatic compaction. The documentation includes installation commands for profiles, restart verification steps, environment variable switches, and limitations such as resync considerations when used concurrently with billion.
Read MoreStory-Short-Scan: Introduction to the Short Web Novel Ranking Scanning Skill
story-short-scan is a Skill categorized under content-creation, analyzing popular short-story data from platforms such as Zhihu Yanyan, Qimao, Heiyan, and Dianzhong to capture trending themes. Released by an organization located at @org-eyw2ohcx/worldwonderer-oh-story-claudecode-story-short-scan, it is published by RealQ, version 1.0.0, and requires no API key. It can be triggered using /story-short-scan, /短篇扫榜, or phrases such as “短篇什么火” or “知乎故事排行”. This article introduces its positioning, core features, trigger methods, typical use cases, and applicable limitations. The official installation command and the full SKILL.md content could not be verified in the provided materials, and this has been stated as-is. It is recommended to check the official page before use.
Read MoreSimulation Experiments in University Chemistry and Chemical Engineering: Transforming Key Concepts of Chemical Engineering Principles into Online Simulation Links
This is the education skill "Simulation Experiment - University Chemistry & Chemical Engineering" on SkillHub, located at @yimubiao/university-chemeng-sim. Version 1.0.1, published by yimubiao, sourced from enterprise, and requires no API key. It does not possess a built-in simulation engine but instead maps chemical engineering principle keywords mentioned by the user (McCabe-Thiele distillation, absorption tower, vapor-liquid equilibrium P-x-y/T-x-y, Peng-Robinson, Rankine cycle, CSTR reactor, heat exchanger, molecular 3D models, VSEPR) to hyperlinks of online simulation pages with ?kp= parameters. If no match is found, it outputs the skill homepage link. Queries in both Chinese and English can yield results, such as 精馏/McCabe-Thiele/distillation, 朗肯循环/Rankine cycle, CSTR, and VLE. The coverage encompasses online simulations of core equipment in chemical engineering principles: distillation, absorption, heat exchange, Rankine cycle, reactors, and phase equilibrium. The material does not provide specific installation commands or configuration fields; it only lists verifiable page URLs, coordinates, classification, and API key requirements, noting that installation methods should refer to real-time information on the SkillHub page. At the time of retrieval, the page data showed a rating of 100000, 961 downloads, and 0 installations, which are subject to change over time.
Read Moredsh-juzpost: Connecting DSH to JuzPost Social Media Scheduling via MCP
dsh-juzpost is a DeepSeek Harness workflow plugin maintained by JuzPost (MIT, 0.1.1). It connects the DSH web profile to JuzPost MCP, registering the JuzPost service, the JuzPost MCP tools (mcp__juzpost__*), and a posting skill. It supports listing accounts, checking platform limits, and uploading media, as well as drafting, scheduling, or publishing to platforms such as YouTube, Instagram, Facebook, TikTok, X, and LinkedIn. Installation: dsh plugin --profile web add dsh-juzpost; requires JUZPOST_API_KEY (a JuzPost paid-plan API key).
Read MoreUniversity Counselor Work Record Generator: Organizing Fragmented Tasks into Standardized Records
The Counselor Work Log Generator (@user_28dd1a41/counselor-work-log) is an "education" category skill on SkillHub, published by community developer user_28dd1a41. Versioned at 1.0.0, it requires no API Key. Designed to address the pain points of daily administrative record-keeping for college counselors, it recognizes intents such as "record work," "work log," "counselor record," "write work content," "work category," "classroom visits," and "heart-to-heart talks." It supports organizing dictated points, photos, or screenshots into standardized record texts organized by work category. However, it is not suitable for non-counselor work records, purely text-based chats, or knowledge Q&A. The article outlines core features, the enabling method via coordinate installation, three types of reproducible use cases (dictation generation, category-based generation, and photo generation), precautions during use, and provides the official URL.
Read MoreNi Haixia's Zhouyi Liu Ren 2 · Yang House and Five Elements: Formalizing Q&A from Ni Haixia's Courses into a Reusable Skill
This article introduces the Community Skill "Ni Haixia Zhouyi Liuren 2: Yangzhai and Five Elements." Based on the transcript of Lecture 2 of Ni Haixia's "Zhouyi Liuren," it covers Yangzhai (residential feng shui) applications (such as the correspondence of names and positions, auspicious and inauspicious positions of the Eight Trigrams relating to family roles, and directions for childbirth) and a continuation of the discussion on the generation and restriction of the Five Elements. It is published by user_96094b40, Version 1.0.0, Category: professional, requires_api_key=false. The text notes that ratings and download counts are based on crawled data. The provided material does not include details regarding the skill's internal prompts or external tool invocation. The installation entry provides only the official address, and the applicable scenario is limited to standardizing the Q&A tone for the course.
Read MoreDX-T1-05 Licensing Negotiation Support: Technology Commercialization Negotiation Skills for Physicians and Pharmaceutical Companies
This article introduces the Agent Skill "DX-T1-05 License Negotiation Support". Released as a professional-category skill by the Baitong AI Service Platform (enterprise) at version 2.2.5, it is located at the coordinate @baitongai/dx-t1-05-license-tan-pan-zhi-chi. Requiring no API Key, it currently holds a rating of 100,000, with 888 downloads and 0 installations. Designed for use when doctors or pharmaceutical companies require the niche service of license negotiation support, it falls under the Technology Transfer T1 New Indication Discovery (Drug Repurposing) coordinate within the P1L3/P1L4 cross-matrix. The article details the pain points it addresses, core functional boundaries, installation and activation methods (noted transparently as no explicit installation commands are provided in the source materials), typical trigger scenarios, as well as its scope of application and limitations.
Read MoreContent Creation Skill: Automated Generation of Construction Logs and Supervision Logs
This article introduces the "Content Creation Skill" (ID: @user_9c258f7c/sdadasf1112, Version 1.0.0, Source: Community, Category: Content Creation) released on the SkillHub community. Based on data from infrastructure, station construction meetings, and construction supervision, this skill automatically generates engineering documents such as construction diaries and supervision logs, requiring no API Key for use. However, the provided documentation contains only a brief one-sentence introduction without specifying template coverage, field mappings, or output formats. With zero installations, it currently lacks real-world usage feedback. Furthermore, conflicting information suggests the skill is intended to remove AI writing traces from text, which contradicts the description regarding engineering document generation. The article relies on SkillHub's scraped data for the introduction and explicitly notes this discrepancy in the precautions section.
Read Moredsh-baidu-netdisk: Integrating Official Baidu Netdisk MCP into DSH
dsh-baidu-netdisk is a community networking tool plugin that registers the SSE tools from the official Baidu Netdisk MCP (baidu-netdisk/mcp) into DeepSeek Harness, requiring users to authorize via OAuth. It provides approximately 13 read-only tools by default, expanding to 23 write operations when `enableWrites` is enabled; the tool prefix is `mcp__baidu_netdisk__`. The installation command is `dsh plugin --profile desktop add github:justinzyc271828-sys/dsh-baidu-netdisk`. Authorization requires Node.js 22+ to run `auth.js`, with tokens stored in `~/.dsh-baidu-netdisk/`. Write calls require DSH approval.
Read Moredsh-ai-jobs: Building an AI Job Radar in DeepSeek Harness
dsh-ai-jobs is a web integration tool plugin maintained by justinzyc271828-sys for DeepSeek Harness. It indexes job postings from approximately 71 AI-related companies by accessing public ATS endpoints such as Greenhouse, Ashby, and Lever. The plugin performs local normalization, rule-based classification, and requirement extraction, scores jobs by skills and location, and supports search, cross-company requirement aggregation, and Markdown daily reports. It requires no API key, and all data is stored locally. Based on the GitHub README, this article introduces installation commands, tool configuration essentials, and usage considerations.
Read MoreNonprofits’ AI Knowledge Steward: Automatically Archive Scattered Files, Build Indexes, and Track To-Dos
This article introduces the SkillHub community skill "Public Welfare Organization AI Knowledge Steward" (coordinates @user_edd42735/insigoo-memory, version 1.0.2, category knowledge-management, requires_api_key=false, community-sourced; page metadata shows a rating of 100000 and 977 downloads). Targeted at knowledge base maintenance scenarios for public welfare organizations, it provides features such as automatic upload classification and archiving, automatic indexing, active detection of schedules/to-dos, and supports knowledge base initialization, file structure scanning, knowledge health checks, and the generation of a 9-zone dashboard. The article explains the installation entry points (official address and SkillHub page) and the five-step usage of a typical diagnostic workflow, while also noting limitations such as the community-origin source, the non-elaboration of judgment rules, and the unspecified format for schedule sources. Verification notes: The body of the official page could not be retrieved at the time of writing (the page is front-end rendered); the body is based solely on script-prefetched data. Installation commands and platform differences were not included; please refer to the official address for details.
Read MoreTencent Lobster Skill Packaging Assistant: Packaging local skills into publishable zips
Introducing the community skill “Tencent Lobster Skill Packaging Assistant” (@indiv-wangxia/skill-publisher, v1.4.0, SkillHub category: ai-agent): it one-click validates and packages locally debugged skills into a zip that meets Tencent Skill Market store publishing specifications, and generates ready-to-copy form fields such as Slug, display name, icon, description, version number, and release notes. The article explains its three verified capabilities, the API keys and editor requirements that need to be confirmed before enabling it (no installation commands are provided in public materials, and the official page failed to be retrieved this time), the workflow from local directories to submitting the store listing form, as well as notes including its community source, rating of 100000, 958 downloads, 0 installs, and that it requires an API Key.
Read MoreMaster of Short Story Writing: A Skill That Hands the Story Over to the "Novelist"
"Short Story Writing Master" is a content-creation Skill released by community developer user_60a201c0 (Skill version 1.3.4). It encapsulates the entire process of short-story conception, writing, and finalization as a callable capability: it takes a story idea as input and produces a short story with a well-rounded structure, vivid characters, and a tightly paced plot, aimed at being ready for direct publication. Usage does not require an additional API key. This article introduces its positioning, core features, installation and activation methods, typical usage, and applicable scenarios, and specifies the data sources and points that require empirical verification.
Read More