Projects | Dylan Chiang

DC-WeMark: Local-first WeChat public account Markdown typesetting tool

Sat, 11 Jul 2026 00:00:00 +0000

DC-WeMark is an open source Markdown typesetting tool for WeChat public account authors. Users can write Markdown on the left, preview it in real time on the right, and paste style-preserving rich text into the official account editor with one click. The entire processing process is completed locally in the browser, and no account, backend or article upload is required.

What I am responsible for in the project

This is an independent open source project designed and developed by me. I am responsible for the complete process from problem definition, technical architecture to product delivery.

Product Design: Focus on format loss, repeated adjustments and content privacy issues when Markdown authors enter the public account editor, and establish a single workflow of “typesetting, previewing, and copying”.
Core Engine: Parse Markdown AST with Rust and comrak, convert theme rules into inline CSS acceptable to public accounts, and compile to WebAssembly.
Front-end experience: Use Vite and native TypeScript to complete dual-column editing, synchronous scrolling, mobile preview, file import and export, draft saving, and clipboard compatibility.
Design System: Create 48 original typography themes and provide adjustable parameters such as font, font size, accent color and copy background.
Project Delivery: Complete Rust unit testing, Playwright end-to-end testing, Docker self-racking solution, and Cloudflare Pages’ continuous deployment process.

Core Innovation

1. Put the typesetting core into WebAssembly

Markdown parsing, theme application and compatibility processing are all completed by the Rust core, and then executed in the browser through WebAssembly. The same core set can be extended to the CLI or desktop in the future without having to re-implement the layout rules.

2. Content does not leave the user’s device

The system has no server logic, database, or account services. Articles are always retained locally from input to copy, taking into account the privacy and offline availability of sensitive content.

3. Compatible output for official account editor

General web page CSS cannot be saved directly in the official account editor. DC-WeMark inlines styles into elements one by one, filters unsupported tags and attributes, and writes them to the clipboard using text/html to make the pasted layout as consistent as possible.

4. Small size, low dependency product architecture

The front end does not rely on a large UI framework, and the deployment artifacts are files that can be hosted by any static service. This keeps the online version, self-hosted version, and future multi-device reuse simple.

Technical architecture

Typesetting Core: Rust, comrak, WebAssembly
Front-end: Vite, TypeScript, HTML, CSS
Quality Assurance: Rust tests, Playwright E2E
Deployment: Cloudflare Pages, Docker, nginx
Authorization: MIT License

BOYA: Humanities and Social Sciences AI Research Skills Library

Fri, 12 Jun 2026 00:00:00 +0000

BOYA is an open source AI paper workflow I designed for researchers in the humanities and social sciences. It breaks tasks such as topic selection, literature retrieval, research design, paper revision, citation checking, oral examination and submission into skills that can be used independently or run in relays, allowing AI to handle the heavy work while leaving research judgment and academic responsibility to the researcher.

What I am responsible for in the project

I am responsible for BOYA’s problem definition, workflow architecture, all skill design, actual test verification and open source maintenance.

Requirements Translation: Convert tacit methodological knowledge in humanities and social science research into questions, checkpoints, and output formats that can be used by researchers with non-programming backgrounds.
Workflow Design: Establish a navigation portal and sixteen dedicated skills, covering research questions, literature, theoretical framework, methods, writing, review, oral examination and submission.
Human-computer collaboration governance: Set up manual decision-making gates for research questions, theoretical frameworks, method selection and interpretive judgments to prevent automation from overtaking researchers.
Academic Integrity Mechanism: Make “no fabrication found”, “references must be made to the source” and “transparent disclosure of AI use” as hard rules in the workflow.
Verification system: Use real paper materials to establish actual test cases, structured evaluations and evidence records, and continuously write down the exposed problems back to the technical rules.
Cross-platform maintenance: Supports Codex, Claude Code and OpenCode, and provides Traditional Chinese, Simplified Chinese, English and Japanese portals.

Core Innovation

1. Convert research methods into executable workflows

BOYA not only provides prompt words, but also breaks down each research stage into clear inputs, judgment gates, outputs and handover conditions, making AI collaboration a traceable research process.

2. Human-in-the-loop Hard Decision Gate

The process can assist with retrieval, sorting, checking and simulated questioning, but when encountering research questions, theory selection, method selection and proposition revision, the researcher must stop and make decisions. This is the core difference between BOYA and fully automatic essay generation tools.

3. Distinguish between bibliographic existence and claim establishment

reference-check is responsible for confirming the literature and bibliographic fields, while claim-audit goes back to the original text to check whether the key claims are supported to avoid mistaking the existence of a DOI as evidence of content.

4. Use real research materials to drive iteration

The boundaries of each skill are exposed through real research cases, and error patterns are written back into rules and evaluations. The reliability of workflow comes from checkable actual measurement records, not just functional descriptions.

Project results

Build 17 composable research skills with complete navigation workflows.
Accumulated 20 real material test cases.
Provide offline unit testing, structured regression cases and cross-model manual verification process.
Open source under MIT License and provides usage entrances in four languages.
It has been used for reference review of the author’s master’s thesis, and errors in DOI, signature and bibliographic information were found and corrections were made public.

WeiShi (未識): AI Relationship Exploration for the Sixth Yunnan-Taiwan University Student Innovation and Entrepreneurship Competition

Sun, 24 May 2026 00:00:00 +0000

WeiShi is an AI relationship-exploration project developed for the Sixth Yunnan-Taiwan University Student Innovation and Entrepreneurship Competition (第六屆雲台大學生雙創賽). Its proposition is simple: AI avatars connect first, while people retain the final decision. The prototype turns an initial encounter from quick browsing and instant judgment into an AI-assisted process of mutual understanding.

What I am responsible for in the project

I am mainly responsible for product and technology, transforming initial ideas into complete product prototypes that can be actually operated and displayed offline.

Product Definition: Establish the core proposition of “knowing before seeing” and position AI as a medium for understanding before meeting, rather than a chatbot that makes relationship decisions for users.
System Design: Planning the complete process of avatar profile creation, Agent Plaza, weekly in-depth recommendations, pre-understanding reports and real-person takeover.
Prototype Implementation: Connect the main interactions and local status in series to complete a front-end product that can run offline on the competition computer.
Governance Implementation: Transform the principles of authorization, observability, non-scoring and local priority into actual product nodes and operational restrictions.
Result Integration: String products, scenarios, governance research and roadshow narratives into a verifiable minimum closed loop, so that innovation is not just a concept.

Project Concept

Wei Shi takes “knowing before seeing” as its core concept. Users first create and continuously calibrate their own AI avatars, and then the two avatars conduct preliminary communications around life goals, relationship expectations, communication methods, and important differences. Only after both parties authorize and confirm their understanding of the report, the system will open the anonymous text conversation with real people.

The following diagrams are original Chinese-language competition materials retained as historical evidence; the English captions and context on this page are provided for international readers.

Core process

Create AI as me avatar: The user corrects memory, values, relationship goals and interaction boundaries.
Enter the Agent Square: Browse the content shared by other avatars and form a more complete understanding before formal contact.
Weekly in-depth recommendation: The system provides an in-depth object, presenting the reasons for recommendation, differences to be confirmed, and verification status.
Generate pre-understanding report: Organize the agreement and differences between the two parties on multiple issues, and retain questions that require real people to answer.
Live person takes over the conversation: After both parties agree, switch from agent interaction to anonymous text communication.

Core competitiveness and innovation

1. From chat tools to relationship understanding protocols

Weishi does not pursue more and faster matching, but allows both agents to complete authorizable and reviewable pre-understanding before meeting in person. The value of AI is not to decide relationships for people, but to reduce ineffective chats and information gaps.

2. In-depth rhythm of one person per week

The product replaces unlimited browsing with “one deep subject per week”. After multiple rounds of Agent communication, each object forms an understanding report, allowing users to see specific consistencies, differences, and issues to be confirmed.

3. AI clone that is observable and cannot exceed authority

The avatar can only express within the scope of the user’s authorization; the communication process can be reviewed, and sensitive actions need to be confirmed again. The real person always reserves the right to make the final decision.

4. Don’t judge relationships by a single score

The understanding report presents evidence, discrepancies and real-life takeover issues on multiple topics, without using an overall score to simplify complex relationships into “suitable” or “unsuitable”.

5. Turn governance into a product mechanism

Local processing, authorized nodes, process traces, and non-repudiation commitments are not additional terms, but product capabilities directly written into the interactive process.

Design principles

Human-in-the-loop: AI assists in organizing and understanding, but does not make promises, define relationships, or decide to meet for the user.
Consent takes priority: clear authorization nodes are reserved for pre-contact, report confirmation and real-person takeover.
Local First: Competition prototypes can be run offline, and personal correction status is saved locally in the browser.
Understand not score: The report presents specific issues and evidence, and does not simplify the relationship with a single matching score.

The current version is a local front-end prototype used for competition display and product verification. It does not include a formal account, online model service, real matching algorithm or production data platform.

Home task management system: Cloudflare full-stack collaboration platform

Thu, 02 Oct 2025 00:00:00 +0000

The family task management system is an open source collaboration platform for families and small teams. It integrates tasks, calendars, inspiration boxes, periodic schedules, member divisions and notifications into the same interface, and uses Cloudflare’s edge services to complete front-end and data layer deployment.

What I am responsible for in the project

This is a full-stack open source project designed, developed and maintained by me. I completed a complete refactoring from an early PHP version to an end-to-end TypeScript architecture.

Product Definition: Starting from the needs of family members to jointly manage daily tasks, core processes such as workbench, tasks, calendar, inspiration box, classification, member and personal settings are designed.
Architecture Refactoring: Refactor the PHP and MySQL monolithic systems into an edge-native architecture composed of React, Hono, Cloudflare Workers, D1, KV and R2.
Data and API: Design data models, validation rules and APIs for teams, tasks, categories, comments, notifications, periodic tasks and schedule blocks.
Complex task logic: Implement periodic task generation, time window, backlog tasks, calendar view and schedule update, and handle cross-layer data consistency.
Identity and Permissions: Create JWT access tokens, KV refresh sessions, team isolation and role permissions.
Engineering quality: using monorepo, front-end and back-end shared Zod schema, type checking, unit testing and Cloudflare automatic deployment.

Core Innovation

1. Shift from personal to-do to family collaboration

The system not only manages the list of a single user, but also supports member assignment, classification, comments, notifications and multi-view workbench based on family or small team units.

2. Unify multiple tasks with time semantics

General tasks, periodic tasks, time windows, itineraries and inspiration boxes each have different life cycles. The system allows different tasks to be managed in a consistent operating model by sharing data contracts and calendar views.

3. End-to-end TypeScript contract

The front-end, back-end and shared packages are located in the same pnpm workspace, and Zod schema and type sharing are used to reduce API field drift, allowing input validation and type safety to extend from the interface to the data layer.

4. Edge-native low-dimensional operation architecture

The frontend is deployed on Cloudflare Pages, the API runs on Workers, and the data uses D1, KV and R2. This architecture reduces traditional server maintenance while retaining database, cache, file and scheduling capabilities.

Technical architecture

Front-end: React, TypeScript, Vite, Tailwind CSS, shadcn/ui, PWA
Backend: Hono, Cloudflare Workers, Zod
Data layer: Cloudflare D1, Drizzle ORM, KV, R2
Engineering: pnpm workspace, Vitest, MSW, Cloudflare Pages
Authorization: MIT License

Kinmen AI Computing Center (金門 AI 算力中心): A Project for the Fourth Yunnan-Taiwan University Student Innovation and Entrepreneurship Competition

Sat, 20 Jan 2024 00:00:00 +0000

The Kinmen AI Computing Center is the computing-infrastructure concept I proposed for the Fourth Yunnan-Taiwan University Student Innovation and Entrepreneurship Competition (第四屆雲台大學生雙創賽), where it received a Silver Award. The project uses Kinmen’s regional location as a starting point to study how GPU compute, cloud services, technical support, and compliance governance can form an implementable service model.

This is a research and competition proposal, not an operating physical data center. Its value lies in placing computing demand, regional development, business models, and technology governance within the same feasibility framework. Any visual materials and linked source documents are retained in their original Chinese as historical competition evidence.

What I am responsible for in the project

I am mainly responsible for the overall concept, research analysis and integration of technology and business models.

Problem Definition: Starting from the computing power threshold of AI model training and inference, define the cost and elasticity issues faced by enterprises, research institutions and developers when obtaining GPU resources.
Service Architecture: Planning service modules such as on-demand computing power, long-term leasing, pre-configured development environment, data processing and technical support.
Regional Model: Analyze the geography, industry and cooperation conditions of Kinmen as a regional computing node, as well as the talent cultivation and digital infrastructure benefits it may bring.
Business Design: Establish target customer groups, pricing methods, cooperation channels, resource allocation and staged development assumptions.
Risk Governance: Incorporate export control, data security, intellectual property, regulatory compliance, capital and operational risks to avoid discussing only technology supply and ignoring governance boundaries.
Result Transformation: Organize research content into competition proposals and extended papers, so that entrepreneurial ideas have both academic and policy foundations that can be discussed.

Core competitiveness and innovation

1. Convert high-cost hardware into flexible services

Through on-demand use and hierarchical leasing, users can obtain computing power according to model scale, project cycle and budget, reducing the one-time investment and maintenance costs required to build their own GPU servers.

2. Not just rent GPUs, but provide a complete working environment

The project treats computing resources, model development environment, data processing, performance optimization and technical support as one set of services, shortening the preparation time for users from obtaining hardware to starting work.

3. Connect research and industrial needs through regional nodes

Kinmen is not only a location option, but also a regional innovation node in the project. The concept considers research institutions, enterprises, talent cultivation and local industries at the same time, so that computing infrastructure can form a broader application network.

4. Build compliance and security into the architecture

Computing power services involve data, models, hardware and cross-border specifications. Data isolation, access control, intellectual property rights, export controls and legal compliance are conditions for the project from the outset, rather than added as an afterthought.

5. Connect research, policy and entrepreneurship verification

This project is not just a business plan, but also extends to the study of AI hardware collaboration, regional scientific and technological cooperation and governance risks, so that technical ideas can be tested at three levels: business, policy and academic.

Service concept

Elastic GPU computing power: Configure different levels of computing resources based on hourly, project or long-term plans.
Pre-configured development environment: Provides common AI frameworks and tools to reduce deployment and environment setting costs.
Data and model support: Covers data pre-processing, model deployment, performance optimization and technical consulting.
Security and Isolation Mechanism: Plan data isolation, permission management and backup for different customers and workloads.
Industry-university cooperation scenario: Support research projects, talent training, enterprise PoC and regional digital transformation.

Research and Results

Received a Silver Award at the Fourth Yunnan-Taiwan University Student Innovation and Entrepreneurship Competition.
The extended research is “ ”.
Establish a complete proposal framework from market demand, service architecture, business model to risk governance.

All cross-regional service concepts of this project are based on applicable export controls, data protection and related laws, and do not advocate or design any transactions or technical paths to circumvent supervision.