Skip to main content
ClaudeWave
Skill570 repo starsupdated 2mo ago

light-software-copyright

>-

Install in Claude Code
Copy
git clone --depth 1 https://github.com/Light0305/Light-skills /tmp/light-software-copyright && cp -r /tmp/light-software-copyright/skills/light-software-copyright ~/.claude/skills/light-software-copyright
Then start a new Claude Code session; the skill loads automatically.

SKILL.md

# Software copyright materials

Prepare auditable China software copyright material drafts from a real project.
The core job is not “write pretty documents”; it is to keep software name,
version, source-code material, manual, screenshots, applicant fields and user
confirmations consistent enough for the user or professional service provider
to review.

Read
[`references/software-copyright-resource-map.md`](references/software-copyright-resource-map.md)
before current-rule or material-format work.
Read
[`references/materials-workflow.md`](references/materials-workflow.md)
before source-code selection, business-understanding, rights-basis,
third-party-code or formal-output work.

## Non-negotiable boundaries

1. This skill is not legal advice and never submits applications or claims the
   material will pass. It prepares local drafts and review evidence only.
2. Code material must come from real project source files. Never invent,
   rewrite, pad, or AI-generate source code for deposit.
3. Do not expose secrets. Run or request a secret/privacy scan before exporting
   source-code material; if secrets are present, stop for redaction strategy.
4. Preserve official-rule uncertainty. If the current China Copyright
   Protection Center workflow, form, fee, page format or exception rule is not
   verified that day, write `UNKNOWN` or `PLANNED`.
5. Keep every formal material version and software name consistent with the
   application worksheet.
6. Keep this skill off the research DAG. It emits no `light.findings.v1`, has
   no stage, no checkpoint gate and no reroute.

## Workflow

### 1. Intake and scope

Capture or mark `UNKNOWN`:

- jurisdiction: default `CN`; do not silently reuse CN rules for US/EU work;
- software full name, abbreviation if any, version, owner/applicant, developer,
  completion date, publication status and development mode;
- whether the project is independent, commissioned, cooperative, inherited,
  derivative or based on licensed third-party code;
- target output language, local folder for drafts, and whether screenshots are
  needed.

Stop for user confirmation before formal documents are generated.

### 2. Bind to a real source snapshot and deposit plan

Inventory source files from the project the user placed in scope. Record
relative path, SHA-256 and why selected. Exclude:

- generated build artifacts, minified bundles, vendored dependencies and lock
  files unless the user explicitly says they are the software expression to
  deposit;
- secrets, tokens, keys, credentials, private data and unrelated third-party
  code;
- arbitrary middle snippets chosen only to make the material look longer.

If the project is too small, submit all real source rather than padding.

Run the lightweight planner before extraction:

```bash
python scripts/source_deposit_plan.py --root <project-root> --out 软件著作权申请资料/草稿/source-deposit-plan.json --software-name "<软件全称>" --version "<版本号>"
python scripts/source_deposit_plan.py --selftest
```

Use the plan to confirm candidate files, page mode, selected file count,
secret-scan status and the plan SHA. If secret-like content is found, stop for
redaction before exporting code material.

### 3. Confirm application fields and business understanding

Generate a worksheet for the user to verify:

- software name and version;
- owner/applicant and contributor facts;
- development environment and operating environment;
- main functions, technical characteristics and use scenario;
- source program quantity and document material quantity;
- publication status and dates.

The manual must describe how a reviewer/user operates the software. It should
not be a generic feature list disconnected from screens, commands or workflows.
Record rights basis and third-party/open-source/generated-code review before
formal export. Do not assume independent ownership when the project may be
commissioned, cooperative, inherited, derivative or license-constrained.

### 4. Prepare code and document material

For normal China material logic, follow the official first/last rule:

- source program and one documentation material use front/back continuous 30
  pages when the whole material is at least 60 pages;
- if the whole material is fewer than 60 pages, include the whole source or
  whole document;
- full source pages normally need at least 50 lines, and full document pages at
  least 30 lines, unless a documented official exception applies.

Keep extraction manifests and page counts. If exceptional deposit is requested,
mark it explicitly and ask the user to verify the current official rule.

### 5. Generate local drafts, then formal outputs

Use a local folder such as `软件著作权申请资料/`:

- `草稿/`: application worksheet, manual draft, code-selection notes;
- `正式资料/`: final DOCX/TXT/PDF-like outputs only after confirmation;
- manifest/report JSON: file paths, SHA-256, source snapshot, confirmation
  basis and rule status.

Screenshots are optional; if not used, record why. If used, bind each screenshot
to a real running UI or user-provided image and hash it.

### 6. Run the machine gate before saying “ready for review”

Create a packet following
[`templates/software-copyright-materials.example.json`](templates/software-copyright-materials.example.json),
then run:

```bash
python scripts/source_deposit_plan.py --selftest
python scripts/materials_gate.py --packet software-copyright-materials.json --base <project-root> --as-of 2026-07-05
python scripts/materials_gate.py --selftest
```

The gate must pass before calling the package ready for user/professional
review. A failed gate means repair the source plan, source binding, rights
basis, third-party review, confirmation, page rule, version consistency, secret
handling or output hashes.

## ACT / ASK / NEVER

ACT:

- use real source files and record SHA-256;
- ask the user to confirm application fields, business context, code selection,
  drafts and final export;
- keep source-code and documentation page counts explicit;
- make s
light-backend-codingSkill

后端代码编写、逻辑强、安全性高、可读性好、版本控制、代码审查。当任务需要写实验代码、模型代码、数据处理代码、可视化代码、后端接口或系统逻辑时使用。要求逻辑清晰、安全、可读、可维护、便于复现/扩展/部署。支持 Git 版本管理、代码审查、注释规范、README、依赖管理、环境配置、运行说明与项目结构整理。

light-citationSkill

Verify scholarly references and claim-citation support for Light stage 10. Use when auditing a manuscript, claim map, bibliography, DOI/arXiv/PMID/ISBN/URL, BibTeX/CSL, citekeys, chimeric or fabricated citations, retraction/correction alerts, or preparing a canonical citation registry for typesetting. Builds provenance-preserving inventories, confirms metadata with independent authoritative sources, distinguishes CONFIRMED/CONFIRMED-MISSING/UNAVAILABLE/UNRESOLVED, records Crossref update direction, and emits the citation gate plus delivery artifacts.

light-competitionSkill

竞赛与项目申报材料辅助。当用户做统计建模、数学建模、互联网+、挑战杯、大创、创新创业、科研训练等项目时使用。辅助写申报书、项目计划书、商业计划书、路演 PPT、答辩稿、项目摘要、技术路线、创新点、可行性分析、市场分析、研究基础、预期成果、经费预算、团队分工。用于非论文投稿场景,可与论文/软著/专利/PPT 联动。

light-consistencySkill

>-

light-data-engineeringSkill

>-

light-figure-drawingSkill

从顶会大牛角度进行专业绘图与组图。当用户需要把规划好的图实际画出来时使用。按情况用 Python(matplotlib/seaborn/plotly/altair)、R(ggplot2)、MATLAB、Visio、Origin、LaTeX/TikZ、Illustrator、PowerPoint 等。审美统一、专业清晰、配色合理、字体规范、线条清楚、高分辨率,适合直接投稿。不仅画图,还从论文表达角度判断怎么排、怎么组、怎么标注、怎么突出重点。

light-figure-planningSkill

根据论文内容规划应该做哪些图、哪些表、插在哪里、各起什么作用。当用户需要论文图表规划时使用。图表不限于统计图,也包括数据集真实效果图、模型输出示例、案例展示、可解释性可视化等。规划框架图、技术路线图、数据集示意图、模型结构图、算法流程图、结果对比/消融/敏感性图、真实效果图、统计表/对比表等,以审稿人标准判断哪些必做、哪些冗余。

light-file-readingSkill

>-