liang sun
Quantitative Topic Selection
Turn a vague research interest into a testable, feasible quantitative paper topic tailored to your target journal. Whether you plan to submit to a CSSCI or SSCI journal, or are comparing cross-sectional and longitudinal research, this Skill helps clarify your research direction, population, and scope while building a clear variable framework around independent, dependent, mediating, and moderating variables. As your topic takes shape, you can develop research questions, a theoretical framework, formal research hypotheses, measurement instruments, sampling and data collection plans, and statistical analysis approaches aligned with your hypotheses. If you already have a policy project or research keywords, you can use them to identify directions with academic publication potential. If you have not yet settled on a specific angle, you can generate and compare multiple candidate topics. The final output focuses on topic feasibility and publication potential: whether the data can be obtained, whether the variables can be operationalized, whether the theory, hypotheses, and methods form a coherent chain, whether the target journal is a good fit, and where the topic can occupy a distinct niche relative to existing research. You will receive a complete topic finalization plan covering the research direction, variable relationships, theoretical hypotheses, cross-sectional or longitudinal design, data and analysis plan, risk considerations, and feasibility recommendations—ready to share with your advisor or use as the basis for a literature review.
Teaching AI Designer
An AI-powered teaching design system for classroom teachers, combining two modes. Mode A uses the OOAT four-question model to turn a specific teaching frustration into a ready-to-paste teaching agent with an identity, rules, and five seeds. Mode B follows the complete lesson-planning workflow: context injection → in-depth material review → lesson-planning package → formal lesson plan. It includes a six-part framework for teaching frustrations, three-trap and dual-core audits, and compliance guardrails for humanlike interaction. Platform-neutral, with outputs that can be used on any large language model or agent platform.
Course Competency Mapper v2.0
An automated system for generating higher vocational course competency maps aligned with Document No. 1 of Vocational Education (Jiaozhicheng [2026] No. 1). Following the reverse design logic of 'Position → Task → Competency → Skill Point → Course → Assessment', it guides through six stages to produce review-level competency maps: position anchors, typical work tasks, five-level structure breakdown, full-chain mapping matrix of position-competency-course-assessment-certification, quantified passing lines, credit hour validation, and red-line self-checks. Built-in anti-hallucination firewall (three-option source tags, no fabrication, source coverage self-check) and parameterized ideological-political mapping (skill points → professional qualities → observable assessment points), automatically handle conflicts in credit hour standards from multiple materials and consolidate with talent cultivation plans.
IP Integration Course Builder
Transform a course syllabus into a practical IP integration model. The assistant refines the integration logic based on course content and professional training goals, then names the model with a meaningful, discipline-specific title. It provides a core framework, dynamic implementation path, and three-dimensional verification indicators to ensure IP integration is aligned with the professional core and vocational skills. Three optional variants: abbreviation naming (using discipline-specific acronyms), a three-element version (professional root, ability vein, IP soul), and model visualization (embedded SVG diagram). Includes a five-stage implementation framework (cognitive integration → knowledge immersion → ability refinement → practice deepening → accomplishment sublimation) and three-dimensional indicators (value shaping, knowledge/ability, behavioral transformation).
ResearchMulti-Agent: A-Share Pick & IC
It's not an AI assistant, but a virtual investment research team. Common AI stock-picking tools suffer from three problems: fabricating financial figures and target prices, giving vague "bullish/bearish" remarks, and offering "buy" recommendations without clear reasoning. The Multi-Agent Investment Research Team tackles these with a three-pronged approach: 6 parallel roles, cross-validation, and mandatory source attribution. It convenes researchers, fundamental analysts, technical analysts, sentiment analysts, risk officers, and investment managers to work in parallel, deliberating like a real investment committee. What you get is not fuzzy opinions, but a professional research document with facts, signals, disagreements, risks, and every number traceable to its source. Two modes covering "researching a single stock" and "screening a batch of stocks" Mode A: Single-Stock Committee Deep Analysis — Just provide a stock (e.g., "Analyze BYD 002594"), and the skill automatically convenes a full investment committee: the researcher aggregates market data, financial reports, research reports, and industry chain positioning, presenting only objective facts; the fundamental analyst issues a financial health scorecard, key changes in the three financial statements, and PEG valuation; the technical analyst evaluates trends, moving averages, MACD, support and resistance levels, and provides a five-point buy signal hit table; the sentiment analyst scans institutional divergence, retail investor sentiment, and potential misinterpretations; the risk officer digs up counter-evidence, systematically refuting optimistic conclusions from other roles; finally, the investment manager, without adding new data, integrates everything to produce committee minutes and a one-page summary. Mode B: Multi-Condition Stock Screening — From a specified universe (e.g., CSI 300, a sector/theme basket, or your own stock pool), apply a three-layer funnel: L1 financial hard screen (three consecutive quarters of growth, ample cash flow, PEG<1 or huge increase in contract liabilities), L2 technical timing (base breakout, moving average golden cross, volume breakout, strong pullback on low volume, MACD crossing above zero line), L3 information validation (research report ratings and industry chain logic, eliminating "pure technical without fundamental basis" picks). After obtaining a candidate list, the top N stocks can automatically proceed to Mode A for deep analysis. What you will get Mode A delivers a fixed "five-piece set": ① Full analysis report integrating all six roles; ② Data source and evidence table, with each key conclusion mapped to "data → source → date"; ③ Meeting-style committee minutes (agenda → each role's view → disagreements → consensus → variables to track); ④ Risk list sorted by high/medium/low severity; ⑤ One-page investment manager summary condensing core logic, key variables, verification points, and confidence level. Mode B delivers: Candidate stock list table (ticker | name | triggered conditions | key data | source | trigger date) plus screening criteria and methodology description, optionally with the full five-piece set for top candidates. All outputs are saved as files with ticker and date in the filename for easy reuse and archiving.
Professional Course Topology Diagram and Curriculum Program Material Generator
Input a professional course schedule, and the system automatically identifies competency lines and prerequisite relationships, generates a high-definition topology diagram using code, and produces training program description materials; it is fact-anchored with zero fabrication, and the diagram is created only after verification. Input the professional course offerings in any university's talent training program (PDF/image/text table are all acceptable), and the system automatically identifies the topological relationships between courses—grouping them by professional competency lines, identifying strong prerequisite dependencies and competency responses, locating core hub nodes, handling limited elective/specialized courses, and connecting them to the internship in the 5th/6th semester; then, it uses HTML/SVG code to accurately render a high-definition course topology diagram with clear Chinese characters and unblurred lines (not an AI-generated diagram), and finally produces a talent training program curriculum system description material that conforms to the style of teaching documents and includes diagrams. Built-in fact-anchoring and manual verification gate control: course names/nature/class hours are strictly faithful to the original table with zero fabrication; prerequisite relationships are clearly marked as teaching logical inferences and are verified by the user before diagram creation; information such as unfounded curriculum standard adjustments is never fabricated.
Book2Skill — Distill Any Book
Stop taking notes — turn books into callable skills. Turn any book into a set of skills you can actually call inside YouMind. An eight-stage pipeline reads the whole book, extracts its methods, stress-tests them, and registers each as a one-click skill — every one carrying its source, triggers, and boundaries. Distill, never fabricate. Stop collecting reading notes; put the book to work.
ResearchAI Master Clone Distillation
Clone any master into an AI personality you can talk to. 4 features: (1) Auto-collect material → Pain point: Scouring the web for content, copy-pasting one by one (2) Read and archive each piece → Pain point: "It sounds convincing but says nothing real" (3) Keep original quotes verbatim → Pain point: "It makes up convincing nonsense" (4) One-click assembly into a Skill → Pain point: No technical knowledge, no idea how to implement How to use: Just say "Create an AI version of XXX [e.g., Charlie Munger]" and it handles the rest—you only need to confirm at a few key steps. Social proof: Fully tested with Charlie Munger—6 public articles distilled into 25 psychological bias checklists, 32 verbatim quotes, and a conversational AI Munger... Five red lines are hardcoded into the process to ensure "distillation, not fabrication."
ImagePortrait Freedom-2-Step Prompt
Generate a high-end portrait that is uniquely yours from a single reference image. Many people creating AI portraits upload a selfie and say 'Generate a high-end professional headshot for me,' but often end up with a face that doesn't look like them, a style that is template-like and generic, plastic-looking skin, and makeup and clothing that don't match their personality. The root problem is asking the AI to simultaneously guess 'what you look like' and 'what style to use'—it fails at both. This skill breaks it down into two stable steps: First, you provide a single reference image (professional headshot, magazine portrait, black-and-white film, modern Chinese style, etc.). Instead of immediately generating an image, the AI deconstructs its 'photographic language' like a photographer—analyzing lighting, composition, focal length, background, color palette, mood, and texture—and then writes two complete prompt sets for male and female versions separately. Clothing, makeup, hairstyle, and pose are all written differently by gender, fundamentally avoiding disasters like 'putting a man in a silk camisole.' Second, you choose your set of prompts and pair it with a clear photo of yourself: the reference image handles the style, while your photo handles the identity—preserving your face shape, features, age perception, and expression, then applying the professional photographic quality from the reference. The result is a portrait that is unmistakably you with a cinematic quality. The reference image controls the style, your photo controls the identity—it's that simple. Changing the reference image changes the direction. ID photos, professional headshots, magazine covers, black-and-white film—one workflow handles them all, completely eliminating the homogeneity and plastic feel of AI portraits. Target audience: professionals needing a headshot/ID photo/personal brand portrait, social media influencers, job seekers, and anyone who wants a high-end portrait that is truly their own.
8-Expert Editorial: AI Writing
An AI-powered collaborative writing system for self-media bloggers, writers, educators, and researchers. It consists of a pipeline team of 8 virtual experts (Style Evaluator, Topic Planner, Research Gatherer, Knowledge Manager, Lead Writer, Fact Checker, Review Editor, Layout Designer) who collaborate throughout the process from style definition to final layout, producing high-quality content that is in-depth, logically rigorous, and stylistically appropriate.
ResearchCN Patent Disclosure Generator
From project documents to deliverable technical disclosure: full process of patent point mining, online novelty search (prioritizing CNIPA + Google Patents/Scholar dual channel), desensitization drafting, and self-check closure. Adapted to YouMind platform toolchain, automatically scanning kanban materials, outputting disclosure document + independent novelty search report, supporting generateDiagram diagram generation, iterative revision, and version management. Built-in P1-P7 core protocols (factual integrity/auto desensitization/version protection/confirmation gate/self-check isolation/novelty search positioning/social science context).
ResearchNSSF Topic Incubation
Helps young humanities and social science teachers applying for the National Social Science Fund for the first time. Starting from vague research directions, it uses a four-stage dialogue incubation process (direction analysis → theory reconstruction → argument draft → self-check optimization) to ultimately produce a high-quality topic and argument draft ready for submission.
Tutorial & Handbook Workflow
A one-stop book generation engine from research to finished book. Supports deep research, intelligent outlines, AI illustration generation, and professional typesetting, finally outputting complete book documents or web pages. Ideal for creating knowledge manuals, tutorials, booklets, and other content products.
ResearchDistill a Book into Skills
Distill a book into a set of executable skills that can be directly invoked in YouMind. Eight-stage serial pipeline: full book reading → five types of extraction → triple verification → RIA++ specification → relationship networking → red team stress testing → one-click registration. Five red lines lock in 'distillation not fabrication'—each skill comes with original source, trigger conditions, and usage boundaries. Say goodbye to fragmented reading notes, and truly integrate a book into your workflow.
ImageCity Collector Posters · LIANG
Based on the 'NAICHA DESIGN · Collector's Poster' creation standards, translates city/object/abstract themes into exhibition-level posters with oil painting impasto texture, master-level serif typography, diagonal comprehensive color band main structure, and hot-stamping material large text. v1.1 iteration: Enhanced main visual anchor + three-level font hierarchy + Chinese main text with pinyin subtitle.
Voc. Thesis Rev.: E-com & Live
A thesis review skill designed specifically for vocational college majors in network marketing and live streaming e-commerce. From the perspective of an associate professor tutor, it conducts an in-depth review in 5 steps: Form Compliance → Topic Selection & Abstract → Main Structure → Argumentation & Data → Academic Standards & Overall Evaluation. At each step, it produces a structured scoring sub-table and a list of tutor comments, finally culminating in a complete scoring summary table and a warm tutor letter-style feedback. It incorporates the 2026 Ministry of Education new national standard (GB/T 7713.1—2026), plagiarism check ≤30%, AIGC ≤25% compliance benchmarks, and accurately identifies errors in professional practice points such as live streaming product selection, scripts, ad delivery, GMV, and private domain. Suitable for staged review of first draft, second draft, and final draft.