patent-aid: Skill for Claude Code

.claude/skills/idea-to-disclosure/SKILL.md

idea-to-disclosure is a skill for Claude Code from KumnXi/patent-aid. It costs 97 tokens per session (1,094 once invoked), scanned A, original, MIT.

A three-stage workflow that turns a technical idea into a Chinese patent disclosure document in Word format. It can organize the problem, technical solution, and expected effects, then run checks on claims, consistency, unsupported content, and document output.

In plain words
What is it for?
Use it to draft a patent disclosure from an idea, title, technical field, purpose, method, and known shortcomings, optionally including images, validation, history storage, and DOCX export.
Why use it?
It gives an incomplete idea a structured path to a draft while checking for information that was not provided or cannot be supported. It can also use images such as sketches or system diagrams as input.

Skill for Claude Code

Written for Claude Code: user-invocable in frontmatter.

This is KumnXi/patent-aid's own configuration. It tells Claude Code how to work on patent-aid itself, so it is not a mod to install elsewhere. Copy it as a starting point and replace the rules that are about this project. Everything patent-aid configures →

Part of the patent-aid plugin — 6 skills, 1 hook, 3 MCP servers shipped together

Reuse

Borrowing it

Nothing to install: this file belongs to KumnXi/patent-aid. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file
curl -O https://raw.githubusercontent.com/KumnXi/patent-aid/main/.claude/skills/idea-to-disclosure/SKILL.md
Clone the repo
git clone --depth 1 https://github.com/KumnXi/patent-aid

Made for: Claude Code.

Or install patent-aid, the plugin that ships this one along with the rest of its 6 skills, 1 hook, 3 MCP servers.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for idea-to-disclosure

README.md
[![agentmods](https://agentmods.dev/badge/skills/kumnxi/patent-aid/idea-to-disclosure/github.svg)](https://agentmods.dev/skills/kumnxi/patent-aid/idea-to-disclosure)
Your own site
<a href="https://agentmods.dev/skills/kumnxi/patent-aid/idea-to-disclosure"><img src="https://agentmods.dev/badge/skills/kumnxi/patent-aid/idea-to-disclosure/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for idea-to-disclosure

Your own site · 80×15
<a href="https://agentmods.dev/skills/kumnxi/patent-aid/idea-to-disclosure"><img src="https://agentmods.dev/badge/skills/kumnxi/patent-aid/idea-to-disclosure.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 97 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,094 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5.1 $0.00097 $0.01094
Opus 5 $0.00048 $0.00547
Sonnet 5 $0.00019 $0.00219
Haiku 4.5 $0.00010 $0.00109

Measured 12d ago against content hash 7916f36e84c5, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-12, from the pricing page.

Security

Grade A, and why

idea-to-disclosure scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 12d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

.claude/skills/idea-to-disclosure/SKILL.md · 70 lines

How it starts

The opening of the file, as written. The whole thing — 70 lines — stays where its author put it; the contents beside it link to each section on GitHub.

想法 → 技术交底书

项目核心目标:从想法到技术交底书。当用户提供技术想法(而非现成交底书)时,按此流程执行。

输入字段

字段 说明 必需
idea 技术想法描述
title 发明名称("一种...") 建议
tech-field 技术领域 可选
purpose 要解决的问题 可选
core-method 核心方法/技术路线 可选
problems 现有技术不足 可选

想法若含"针对某缺陷…用某方法…达到某效果"结构即可直接生成;缺字段可让用户补充或用默认。

执行步骤

  1. 确认想法已收集:从用户描述中提炼 idea(含缺陷/方法/效果要素)
  2. 若用户贴图(照片/手绘/系统框图/论文截图/PDF):先经 VL"看图"提炼字段,再进入流水线。图片转成结构化字段后才交给 DeepSeek(src/api/vl_client.py 文字桥接):
# 单图或多图:输出 {idea, title, tech_field, purpose, core_method, problems}
D:/Anaconda3/envs/mathmodel/python.exe -m src.utils.idea_vision 图1.png 图2.png
  • 图片内容(设备结构/流程/数值)会被提炼成 idea 字段,缺的字段留空让用户补或走默认
  • 未配置 DashScope / 读图失败时自动回退纯文本输入,不阻断流程
  1. 调用流水线(内部自动执行三阶段 LLM 生成 → 防无中生有修复 → 权利要求校验 → 合规审查 → 加密保存历史 → Word 导出):
D:/Anaconda3/envs/mathmodel/python.exe scripts/generate_patent.py "技术想法" \
    --title "一种..." --tech-field "技术领域" \
    --purpose "要解决的问题" --core-method "核心方法" \
    --problems "现有技术不足" --out output/交底书.docx

3b. 交付自查(可选,交付前建议跑):Word 导出后渲染回看显示质量(公式/附图/布局/乱码),由 VL 看图质检,避免"导出成功但渲染有问题":

# 生成后直接自查;或对已有 docx 单独自查
D:/Anaconda3/envs/mathmodel/python.exe scripts/generate_patent.py ... --selfcheck
D:/Anaconda3/envs/mathmodel/python.exe scripts/selfcheck_disclosure.py output/交底书.docx
  • 原理:docx → LibreOffice 转 PDF → 逐页 PNG → qwen-vl 回看(见 src/utils/disclosure_inspector.py
  • 需 LibreOffice(winget install --id TheDocumentFoundation.LibreOffice);未装时提示安装,不阻断生成
  • 终端显示每页问题数;有 error 级问题建议修复后重新导出

输出检查(生成后核对)

  • 终端应显示:生成模式(llm_staged 为最优)、质检评分(>90 为好)、防无中生有修复数、权利要求校验摘要
  • Word 内含附图:mermaid 流程图经 Graphviz dot 渲染(300dpi),见 src/utils/diagram_generator.py
  • 生成历史已加密保存到 data/disclosure_history/(见 src/utils/history.py
  • --selfcheck 后终端应显示每页自查结论(P01 正常 / 检出 N 个问题)

注意事项

  • 首次运行引擎初始化约 1 分钟(建知识图谱/RAG 索引)
  • 全文约 1.2-1.9 万字,三阶段模式质检平均 96 分
  • 需要真实 DeepSeek API Key(config/api_config.json,保密,不入库)
  • 若用户手上有现成交底书而非想法 → 引导用 /patent-writer 进入撰写阶段

Read the full file on GitHub · 70 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 12d ago First seen · 70 lines · 97 tokens per session scan A 7916f36e84c5

Subscribe to this mod's changes

idea-to-disclosure is a skill published in the GitHub repository KumnXi/patent-aid (2 stars, last pushed 19d ago), licensed MIT. It adds 97 tokens to every session and 1,094 once invoked, about $0.0005 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.

Related

Other skills, from other repositories

patent-disclosure-skill

A Chinese patent-work assistant for finding invention ideas, writing patent disclosures, explaining patents in plain language, tracking policy changes, and preparing responses to examiner notices.

handsomestWei/patent-disclosure-skill · 176 tokens

patent-search

A search tool for published Chinese patent records, using fields such as inventor, applicant, classification number, and title.

handsomestWei/patent-disclosure-skill · 46 tokens

patent-docket

A workflow coordinator for patent case files, from an inventor's technical disclosure to patent application documents. A patent docket is the tracked record of a patent matter, including its stage, questions, drafts, and handoffs.

handsomestWei/patent-disclosure-skill · 93 tokens

patent-application

A patent-application document workflow that turns supplied invention materials into claims, a description, an abstract, and black-and-white drawings. It can also produce Markdown and Word files.

handsomestWei/patent-disclosure-skill · 95 tokens

patent-disclosure

A guided workflow for finding patentable ideas, doing a limited prior-art search, and drafting a Chinese patent disclosure document. A disclosure document describes an invention for later patent drafting.

handsomestWei/patent-disclosure-skill · 40 tokens

patent-oa

An assistant for preparing replies to patent office examination opinions in China. It can use stored past cases and practice manuals to help shape a draft, which must be reviewed before filing.

handsomestWei/patent-disclosure-skill · 43 tokens