Fokkyp/SoftwareCopyright-Skill
中国软件著作权申请材料 生成器 Skills,本 Skills 通过阅读本地项目,自动生成全套 .docx 软著申请材料,全开源,无须再付费购买任何软著申请服务
What it solves
Applying for a Chinese software copyright (软著) is straightforward, but preparing the required documents is tedious: the application form fields must be correct, the operation manual must look professional, and the source code excerpts must follow the official rules (first 30 pages / last 30 pages). Many developers end up paying third-party agencies to do this paperwork, and those services mostly just organize documents. This skill lets developers generate all of these materials locally with their coding agent, without paying a middleman and without sharing code or product details with an external vendor.
How it works
You install the actual skill directory software-copyright-materials/ into your coding agent's skill folder. Then, with your project open in the agent, you trigger the skill (for example, by saying "使用 software-copyright-materials 生成当前项目的软件著作权申请资料"). The skill guides you through the whole flow:
- It analyzes your project to understand the business, pages, and functionality.
- It asks you to confirm key information at several checkpoints: business scope, application form fields, code file selection, screenshot method, and the final Markdown draft.
- It extracts source code only from your real project — it never invents or AI-generates code.
- It automatically applies the official "first 30 pages / last 30 pages" rule for code samples when the source is long enough; if the code is under 60 pages, it generates all of it.
- The operation manual is written from an understanding of your actual product, not from a generic template.
- It collects all application-form fields (software name, version, copyright owner, development environment, runtime environment, source program size, feature description, etc.) into a single
申请表信息.txtthat you can copy into the official online form. - After confirmation, it generates an operation manual DOCX, a code materials DOCX, and the TXT file into
软件著作权申请资料/正式资料/. Everything is generated locally in your project directory.
The skill also performs environment checks at startup, creating 环境检查.md and 环境检查.json that report whether Python 3.10+, OfficeCLI 1.0.151, and optional tools (Playwright CLI, Chrome) are available and configured correctly. If OfficeCLI is missing or mismatched, the skill stops and asks you to install the fixed version or explicitly accept the risk of using another version.
Screenshots can be taken automatically via Playwright CLI (version 0.1.20) if you confirm, or you can place your own images into 软件著作权申请资料/用户截图/. The skill never installs dependencies silently without your consent.
Who it's for
Developers who need to apply for a Chinese software copyright (软著) and want to avoid paying third-party services for document preparation. It's suited for anyone using a coding agent that supports local skill directories, who wants to generate all required materials from their real codebase while keeping full control and privacy.
Highlights
- No cost and no code sharing — generates all materials locally without external agencies.
- Real source‑code extraction — code samples come only from your actual project; AI never fabricates source code.
- Automatic 30/30 rule handling — handles the front‑30 and back‑30 pages rule for long codebases, or generates everything when under 60 pages.
- Business‑aware operation manual — written from understanding your product, not from a generic template.
- Structured application form data — all form fields are collected into a single TXT for easy copy‑paste into the official website.
- Multiple checkpoints — you confirm business details, form fields, code choices, screenshot method, and the final draft before output.
- One‑click output — produces DOCX (operation manual, code materials) and TXT files in a unified folder.
- Full demo included — browse the
生成demo/软件著作权申请资料/folder to see sample drafts and final files. - Environment self‑check — startup checks verify Python 3.10+, OfficeCLI version, and readiness; the skill refuses to fake outputs if dependencies are missing.
Related
- Project
- Project
- Project
- Project
- Project