patentradar: Data & Analytics for AI Agents | OpenAgentSkill

Overview

专利侵权竞品分析 skill。输入专利公开号（CN/US/EP/JP 等），输出可人工复核的 markdown 竞品分析报告。工作流：拆解权利要求 → 搜索市场竞品 → 抓证据（含图像）→ 全部权利要求逐特征对比 → 生成报告。触发词：专利侵权分析、专利竞品、claim chart、专利公开号、CN/US/EP/JP 专利号、专利对比、技术特征对比、专利保护范围、侵权风险评估。不适用：专利申请文书撰写、专利无效宣告、知识产权法律咨询、专利诉讼策略。

Read full documentation

Source documentation, not instructions for this website. Review permissions before running any commands.

PatentRadar 专利侵权竞品分析

⚙️ 执行授权（必读，先于一切）

本 skill 必须以 4 个独立 subagent 串行执行（decompose → competitor_search → full_claim_chart → report），subagent 之间通过 JSON 文件传递数据。

用户已在本 skill 中预先授权你自由派生 subagent，不需要在会话里再次征求确认。 看到本段就视为授权生效，直接按"## 架构"和"## 工作流"开始 spawn。

严禁降级执行：

❌ 不要因为"没有用户明确授权"就退回主对话单轮直答
❌ 不要把 4 个模块合并到一次 LLM 调用里
❌ 不要由主 agent 自己抓页面 / 看图 / 写报告内容
✅ 如果你判断当前环境真的没法 spawn subagent（比如 Task tool 不可用），立即停下来报告原因让用户决定，而不是自作主张降级单轮直答 —— 降级会让证据搜索深度大幅缩水，违反"## 🎯 核心总目标"。

🎯 核心总目标（全流程牢记）

从市场公开渠道找出落入本专利权 1 保护范围的潜在竞品产品，用可追溯的证据（公开 URL + 引文 + 图像）逐特征对比，输出一份能让律师/工程师直接复核的 claim chart 报告，并对仍有缺口的特征明确"还缺什么、下一步可以去哪找"。

理想的产出：找到一篇甚至多篇 total_score ≥ 80（侵权风险阈值）的竞品 —— 这意味着客户拿到报告后能立刻拍板"这家可能在侵权，可以发律师函/起诉了"。如果客观上真存在侵权竞品，搜不到就是失职；但如果确实没有相关侵权的竞品，也不要靠放宽评分标准凑高分。

判断质量的 3 把尺子（任一不达标都算没做到位）：

每个判定都可追溯：明确满足 必须有公开 URL 字面/数值证据（≥ 1 独立 host）；可能满足 必须给严谨推理链；证据不足 / 明确不满足 不能凭印象拍脑袋
数学约束类必须现场算到数值（D/V、S/E、L/S 等），不接受"满足公式约束"这种结论性话术
证据缺口必须可执行：权 1 中所有非"明确满足"的 feature 都必须给出 evidence_gap_brief，让人工拿着它就能去具体网站找具体证据，不要造空泛的"需进一步搜索"

搜索深度与严格评分不矛盾：

搜得全（多语言 query / 垂类站点 / 拆解视频 / 维修手册 / 专利文献 / 规格 PDF / 产品图 / 电路图）是为了别错过真侵权竞品
评分严（不自我宽松、有 ≥1 独立 host 才给"明确满足"、有严谨推理链才给”可能满足“）是为了别给假高分
两者一起做才能稳定挖出真侵权候选

每个 subagent 在执行细节决策时都应先问自己："这一步对达成上述目标有帮助吗？"——避免被局部规则牵着走、忘了大目标。

何时使用

用户给出专利公开号（CN/US/EP/JP 等）并要求做竞品分析 / 侵权风险评估 / claim chart 对比。典型输入：

"分析 CN114512759B 的市场竞品"
"帮我跑下 US10000000B2 的侵权风险"
"对比这个专利和现有产品：[公开号]"

架构

主 agent（你）按顺序 spawn 4 个 subagent，每个 subagent 独立完成一个模块，subagent 之间通过 JSON 文件传递数据。

用户输入专利号 → 主 agent（本 SKILL.md）
                  │
                  ├─ spawn subagent 1 → module_1/task_package.json
                  ├─ spawn subagent 2 → module_2/top_competitors.json
                  ├─ spawn subagent 3 → module_3/full_claim_chart.json
                  └─ spawn subagent 4 → module_4/report.md

每个 subagent 的完整 prompt 在 agents/。

工作流

步骤 0：准备

确认用户给出的专利公开号（如 CN114512759B）。如果不清楚，问用户。

确定输出目录：默认 ./patentradar_output/<PUBLICATION_NO>/，或用户指定的目录。

创建子目录：

<output_dir>/
├── module_1/
├── module_2/
├── module_3/
└── module_4/

步骤 1：spawn subagent — 拆解权利要求

用 Task tool spawn general-purpose subagent，prompt 加载 agents/decompose.md + 用户的专利号。

期望输出：<output_dir>/module_1/task_package.json，符合 schemas/task_package.md 的结构。

跑完后读这个 JSON 验证：

claims 数组非空
主题前序作为首条 feature（如 C1-F1: 一种 XX 系统）

步骤 2：spawn subagent — 竞品搜索 + 权 1 判定

用 Task tool spawn subagent，prompt 加载 agents/competitor_search.md + module_1 输出文件路径。

期望输出：<output_dir>/module_2/top_competitors.json，符合 schemas/top_competitor_report.md。

跑完后验证：

top_competitors 数组，每家公司唯一（同公司去重已做）
每个候选有完整权 1 feature 对比 + total_score

步骤 3：spawn subagent — 全部权利要求扩展

用 Task tool spawn subagent，prompt 加载 agents/full_claim_chart.md + module_1 + module_2 输出文件路径。

期望输出：<output_dir>/module_3/full_claim_chart.json，符合 schemas/full_claim_chart_report.md。

跑完后验证：

每个 TOP 候选有全部权利要求的逐特征对比
权 1 中 status ∈ {可能满足, 证据不足} 的特征都填了 evidence_gap_brief

步骤 4：spawn subagent — 生成报告（markdown + PDF）

用 Task tool spawn subagent，prompt 加载 agents/report.md + module_1 + module_3 输出文件路径。

期望输出：

<output_dir>/module_4/report.md
<output_dir>/module_4/report.pdf（subagent 内部用 WeasyPrint 渲染；若系统库不齐导致渲染失败，允许只有 md）

步骤 5：交付

打印 markdown + PDF 两个路径给用户。同时给一个简短摘要：最高分竞品 + total_score + 是否触发 ≥80 风险阈值 + 失效候选数。

主 agent 的职责（要做什么 / 不要做什么）

要做：

按顺序 spawn 4 个 subagent，每次传清楚 prompt 文件路径 + 输入 JSON 路径 + 输出 JSON 路径
每个 subagent 跑完后，主 agent 必须跑 JSON Schema 校验（用 schemas/validate.py，见下方"校验机制"段）；校验失败时把错误清单透传给该 subagent，让它修正后重交
任一 subagent 都有权随时回查权利要求原文：如果对 task_package 中的 claim_text / feature_text 拆解有疑问，可访问 patent.google_patents_url 或 patent.pdf_url 复核原文（已写进每个 agent prompt 的"能力说明"段）

不要做：

主 agent 不要自己抓页面、不要自己看图、不要自己写报告内容 —— 全部由 subagent 干
不要把 4 个模块合并到一次调用（每个模块用独立 subagent 隔离上下文）

校验机制（每个模块跑完都要做）

每个 subagent 跑完后，主 agent 必须用 schemas/validate.py 跑 JSON Schema 校验，把结果反馈给 subagent（成功就走下一步；失败就把错误清单透传回去让它修正）。

# 模块一跑完后
python skills/patentradar/schemas/validate.py task_package <output_dir>/module_1/task_package.json

# 模块二跑完后
python skills/patentradar/schemas/validate.py top_competitor_report <output_dir>/module_2/top_competitors.json

# 模块三跑完后
python skills/patentradar/schemas/validate.py full_claim_chart_report <output_dir>/module_3/full_claim_chart.json

exit 0 + stdout OK <data_path> → 通过，进下一模块
exit 1 + stdout 错误清单（[<json_path>] <message> 每行一条）→ 失败
- 主 agent 把这份错误清单原样作为消息发回该 subagent："你的输出 schema 校验失败，请按以下错误列表修正后重交：<错误清单>"
- 该 subagent 修正后重写 JSON，主 agent 再校验，最多重试 2 次
- 2 次还过不了，停下来告诉用户哪个字段一直错

注：模块四（markdown 报告）不走 schema 校验——它的输出是 markdown 不是 JSON。主 agent 只检查 report.md 文件存在 + 非空即可。

业务规则速查（4 模块共享，subagent prompt 里也会重申）

业务规则	哪个模块管
权 1 主题前序（"一种 XX 系统"）作为首条 feature	模块一
候选去重 key 用 `(company, product_name)` 二元组	模块二
`product_intro` 是自然语言产品介绍（1-2 句关键参数），不参与去重	模块二
TOP-N 排序时同公司只保留最高分产品	模块二
数学约束类（D/V、S/E、L/S 等）现场算具体数值	模块二/三
`total_score` 只看权 1（非权 1 不进 ranking）	模块三
失效只看权 1（仅权 1 任一"明确不满足"或 launch_date 早于专利申请日才 disqualified）	模块二/三
权 1 缺口特征必填 `evidence_gap_brief`（两行：还缺 / 下一步建议）	模块三

评分规则（模块二/三共享）

status	ratio	触发条件
明确满足	1.0	公开 URL 直接字面/数值证据，≥ 1 独立 host
可能满足	0.8	公开证据严谨推理（必须给严谨的推理链）
证据不足	0.3	证据池里找不到相关线索
明确不满足	0.0	公开证据直接矛盾 → 整候选 disqualified=true, total_score=0

total_score = mean(各 feature ratio) × 100。

垂类网站推荐（模块二/三搜索证据时优先用）

configs/technology_tags.toml 维护两类网站清单（每条带 url + description）：

tags[name == <technology_tag>].recommended_sites：按技术领域的垂类站点（如"动力电池" → batteryfinds.com）
顶层 [[universal_sites]]：所有 tag 通用的高信号站点（车型维修手册站、汽修资料站等）

模块二/三 subagent 在 spawn 时主 agent 应把这个 toml 路径告诉它们，让它们优先在这些站点检索，而不是从零大海捞针。要新增网站推荐时只动这个 toml 即可。

停止条件（subagent 内部判断）

模块二/三 subagent 在搜索阶段：

硬规则：所有权 1 feature 都拿到 ≥1 个独立 host 的明确满足证据 → 停搜
软判断：连续 2-3 轮搜索没找到新有价值证据 → 停搜（subagent 自己评估）
不要无限搜，不要陷入死循环：判断"实在搜不到了"就停，落 证据不足 或 可能满足

触发执行流程

解析用户输入：抽出专利公开号
创建输出目录 + 子目录
顺序 spawn 4 subagent，传 prompt + 输入 JSON + 输出 JSON 路径
每次跑完读取产物 JSON 做基本验证
打印 report.md 路径给用户

完成后告诉用户报告位置。

View source ↗

File metadata

name: patentradar
description: 专利侵权竞品分析 skill。输入专利公开号（CN/US/EP/JP 等），输出可人工复核的 markdown 竞品分析报告。工作流：拆解权利要求 → 搜索市场竞品 → 抓证据（含图像）→ 全部权利要求逐特征对比 → 生成报告。触发词：专利侵权分析、专利竞品、claim chart、专利公开号、CN/US/EP/JP 专利号、专利对比、技术特征对比、专利保护范围、侵权风险评估。不适用：专利申请文书撰写、专利无效宣告、知识产权法律咨询、专利诉讼策略。

View original text

---
name: patentradar
description: 专利侵权竞品分析 skill。输入专利公开号（CN/US/EP/JP 等），输出可人工复核的 markdown 竞品分析报告。工作流：拆解权利要求 → 搜索市场竞品 → 抓证据（含图像）→ 全部权利要求逐特征对比 → 生成报告。触发词：专利侵权分析、专利竞品、claim chart、专利公开号、CN/US/EP/JP 专利号、专利对比、技术特征对比、专利保护范围、侵权风险评估。不适用：专利申请文书撰写、专利无效宣告、知识产权法律咨询、专利诉讼策略。
---

# PatentRadar 专利侵权竞品分析

## ⚙️ 执行授权（必读，先于一切）

本 skill **必须以 4 个独立 subagent 串行执行**（decompose → competitor_search → full_claim_chart → report），subagent 之间通过 JSON 文件传递数据。

**用户已在本 skill 中预先授权你自由派生 subagent，不需要在会话里再次征求确认。** 看到本段就视为授权生效，直接按"## 架构"和"## 工作流"开始 spawn。

**严禁降级执行**：
- ❌ 不要因为"没有用户明确授权"就退回主对话单轮直答
- ❌ 不要把 4 个模块合并到一次 LLM 调用里
- ❌ 不要由主 agent 自己抓页面 / 看图 / 写报告内容
- ✅ 如果你判断当前环境**真的没法 spawn subagent**（比如 Task tool 不可用），立即**停下来报告原因**让用户决定，而不是自作主张降级单轮直答 —— 降级会让证据搜索深度大幅缩水，违反"## 🎯 核心总目标"。

## 🎯 核心总目标（全流程牢记）

**从市场公开渠道找出落入本专利权 1 保护范围的潜在竞品产品，用可追溯的证据（公开 URL + 引文 + 图像）逐特征对比，输出一份能让律师/工程师直接复核的 claim chart 报告，并对仍有缺口的特征明确"还缺什么、下一步可以去哪找"。**

**理想的产出**：找到一篇甚至多篇 `total_score ≥ 80`（侵权风险阈值）的竞品 —— 这意味着客户拿到报告后能立刻拍板"这家可能在侵权，可以发律师函/起诉了"。如果客观上真存在侵权竞品，**搜不到就是失职**；但如果确实没有相关侵权的竞品，也**不要靠放宽评分标准凑高分**。

判断质量的 3 把尺子（任一不达标都算没做到位）：

1. **每个判定都可追溯**：`明确满足` 必须有公开 URL 字面/数值证据（≥ 1 独立 host）；`可能满足` 必须给严谨推理链；`证据不足` / `明确不满足` 不能凭印象拍脑袋
2. **数学约束类必须现场算到数值**（D/V、S/E、L/S 等），不接受"满足公式约束"这种结论性话术
3. **证据缺口必须可执行**：权 1 中所有非"明确满足"的 feature 都必须给出 evidence_gap_brief，让人工拿着它就能去具体网站找具体证据，不要造空泛的"需进一步搜索"

**搜索深度与严格评分不矛盾**：
- 搜得**全**（多语言 query / 垂类站点 / 拆解视频 / 维修手册 / 专利文献 / 规格 PDF / 产品图 / 电路图）是为了**别错过真侵权竞品**
- 评分**严**（不自我宽松、有 ≥1 独立 host 才给"明确满足"、有严谨推理链才给”可能满足“）是为了**别给假高分**
- 两者一起做才能稳定挖出真侵权候选

每个 subagent 在执行细节决策时都应**先问自己**："这一步对达成上述目标有帮助吗？"——避免被局部规则牵着走、忘了大目标。

## 何时使用
用户给出**专利公开号**（CN/US/EP/JP 等）并要求做竞品分析 / 侵权风险评估 / claim chart 对比。典型输入：

- "分析 CN114512759B 的市场竞品"
- "帮我跑下 US10000000B2 的侵权风险"
- "对比这个专利和现有产品：[公开号]"


## 架构

主 agent（你）按顺序 spawn **4 个 subagent**，每个 subagent 独立完成一个模块，subagent 之间通过 JSON 文件传递数据。

```
用户输入专利号 → 主 agent（本 SKILL.md）
                  │
                  ├─ spawn subagent 1 → module_1/task_package.json
                  ├─ spawn subagent 2 → module_2/top_competitors.json
                  ├─ spawn subagent 3 → module_3/full_claim_chart.json
                  └─ spawn subagent 4 → module_4/report.md
```

每个 subagent 的完整 prompt 在 `agents/`。

## 工作流

### 步骤 0：准备

确认用户给出的专利公开号（如 `CN114512759B`）。如果不清楚，问用户。

确定输出目录：默认 `./patentradar_output/<PUBLICATION_NO>/`，或用户指定的目录。

创建子目录：
```
<output_dir>/
├── module_1/
├── module_2/
├── module_3/
└── module_4/
```

### 步骤 1：spawn subagent — 拆解权利要求

用 Task tool spawn `general-purpose` subagent，prompt 加载 `agents/decompose.md` + 用户的专利号。

**期望输出**：`<output_dir>/module_1/task_package.json`，符合 `schemas/task_package.md` 的结构。

跑完后**读这个 JSON 验证**：
- `claims` 数组非空
- 主题前序作为首条 feature（如 `C1-F1: 一种 XX 系统`）

### 步骤 2：spawn subagent — 竞品搜索 + 权 1 判定

用 Task tool spawn subagent，prompt 加载 `agents/competitor_search.md` + module_1 输出文件路径。

**期望输出**：`<output_dir>/module_2/top_competitors.json`，符合 `schemas/top_competitor_report.md`。

跑完后验证：
- `top_competitors` 数组，每家公司唯一（同公司去重已做）
- 每个候选有完整权 1 feature 对比 + total_score

### 步骤 3：spawn subagent — 全部权利要求扩展

用 Task tool spawn subagent，prompt 加载 `agents/full_claim_chart.md` + module_1 + module_2 输出文件路径。

**期望输出**：`<output_dir>/module_3/full_claim_chart.json`，符合 `schemas/full_claim_chart_report.md`。

跑完后验证：
- 每个 TOP 候选有全部权利要求的逐特征对比
- 权 1 中 status ∈ {可能满足, 证据不足} 的特征**都填了 `evidence_gap_brief`**

### 步骤 4：spawn subagent — 生成报告（markdown + PDF）

用 Task tool spawn subagent，prompt 加载 `agents/report.md` + module_1 + module_3 输出文件路径。

**期望输出**：
- `<output_dir>/module_4/report.md`
- `<output_dir>/module_4/report.pdf`（subagent 内部用 WeasyPrint 渲染；若系统库不齐导致渲染失败，允许只有 md）

### 步骤 5：交付

打印 markdown + PDF 两个路径给用户。同时给一个简短摘要：最高分竞品 + total_score + 是否触发 ≥80 风险阈值 + 失效候选数。

## 主 agent 的职责（要做什么 / 不要做什么）

**要做**：
- 按顺序 spawn 4 个 subagent，每次传清楚 prompt 文件路径 + 输入 JSON 路径 + 输出 JSON 路径
- **每个 subagent 跑完后，主 agent 必须跑 JSON Schema 校验**（用 `schemas/validate.py`，见下方"校验机制"段）；校验失败时**把错误清单透传给该 subagent**，让它修正后重交
- **任一 subagent 都有权随时回查权利要求原文**：如果对 task_package 中的 `claim_text` / `feature_text` 拆解有疑问，可访问 `patent.google_patents_url` 或 `patent.pdf_url` 复核原文（已写进每个 agent prompt 的"能力说明"段）

**不要做**：
- 主 agent 不要自己抓页面、不要自己看图、不要自己写报告内容 —— 全部由 subagent 干
- 不要把 4 个模块合并到一次调用（每个模块用独立 subagent 隔离上下文）

## 校验机制（每个模块跑完都要做）

每个 subagent 跑完后，**主 agent 必须用 `schemas/validate.py` 跑 JSON Schema 校验**，把结果反馈给 subagent（成功就走下一步；失败就把错误清单透传回去让它修正）。

```bash
# 模块一跑完后
python skills/patentradar/schemas/validate.py task_package <output_dir>/module_1/task_package.json

# 模块二跑完后
python skills/patentradar/schemas/validate.py top_competitor_report <output_dir>/module_2/top_competitors.json

# 模块三跑完后
python skills/patentradar/schemas/validate.py full_claim_chart_report <output_dir>/module_3/full_claim_chart.json
```

返回：
- **exit 0** + stdout `OK <data_path>` → 通过，进下一模块
- **exit 1** + stdout 错误清单（`[<json_path>] <message>` 每行一条）→ 失败
  - 主 agent 把这份错误清单**原样作为消息**发回该 subagent："你的输出 schema 校验失败，请按以下错误列表修正后重交：<错误清单>"
  - 该 subagent 修正后重写 JSON，主 agent 再校验，最多重试 2 次
  - 2 次还过不了，停下来告诉用户哪个字段一直错

**注**：模块四（markdown 报告）不走 schema 校验——它的输出是 markdown 不是 JSON。主 agent 只检查 `report.md` 文件存在 + 非空即可。

## 业务规则速查（4 模块共享，subagent prompt 里也会重申）

| 业务规则 | 哪个模块管 |
|---|---|
| 权 1 主题前序（"一种 XX 系统"）作为首条 feature | 模块一 |
| 候选去重 key 用 `(company, product_name)` 二元组 | 模块二 |
| `product_intro` 是自然语言产品介绍（1-2 句关键参数），不参与去重 | 模块二 |
| TOP-N 排序时同公司只保留最高分产品 | 模块二 |
| 数学约束类（D/V、S/E、L/S 等）现场算具体数值 | 模块二/三 |
| `total_score` 只看权 1（非权 1 不进 ranking） | 模块三 |
| 失效只看权 1（仅权 1 任一"明确不满足"或 launch_date 早于专利申请日才 disqualified） | 模块二/三 |
| 权 1 缺口特征必填 `evidence_gap_brief`（两行：还缺 / 下一步建议） | 模块三 |


## 评分规则（模块二/三共享）

| status | ratio | 触发条件 |
|---|---|---|
| 明确满足 | 1.0 | 公开 URL 直接字面/数值证据，≥ 1 独立 host |
| 可能满足 | 0.8 | 公开证据严谨推理（必须给严谨的推理链） |
| 证据不足 | 0.3 | 证据池里找不到相关线索 |
| 明确不满足 | 0.0 | 公开证据直接矛盾 → 整候选 disqualified=true, total_score=0 |

`total_score = mean(各 feature ratio) × 100`。

## 垂类网站推荐（模块二/三 搜索证据时优先用）

[`configs/technology_tags.toml`](configs/technology_tags.toml) 维护两类网站清单（每条带 url + description）：

1. `tags[name == <technology_tag>].recommended_sites`：按技术领域的垂类站点（如"动力电池" → `batteryfinds.com`）
2. 顶层 `[[universal_sites]]`：所有 tag 通用的高信号站点（车型维修手册站、汽修资料站等）

模块二/三 subagent 在 spawn 时主 agent 应把这个 toml 路径告诉它们，让它们**优先**在这些站点检索，而不是从零大海捞针。要新增网站推荐时只动这个 toml 即可。

## 停止条件（subagent 内部判断）

模块二/三 subagent 在搜索阶段：
- **硬规则**：所有权 1 feature 都拿到 ≥1 个独立 host 的明确满足证据 → 停搜
- **软判断**：连续 2-3 轮搜索没找到新有价值证据 → 停搜（subagent 自己评估）
- **不要无限搜，不要陷入死循环**：判断"实在搜不到了"就停，落 `证据不足` 或 `可能满足`

## 触发执行流程

1. **解析用户输入**：抽出专利公开号
2. **创建输出目录** + 子目录
3. **顺序 spawn 4 subagent**，传 prompt + 输入 JSON + 输出 JSON 路径
4. **每次跑完读取产物 JSON 做基本验证**
5. **打印 report.md 路径**给用户

完成后告诉用户报告位置。

Use with my agent

Skill source recorded

Skill instructions are recorded. This is not a runtime test, safety guarantee or compatibility certification.

Review before install: Avoid automatic install

License: Unknown

License is unclear
Permission surface may require sandboxing
Repository license is unknown; no explicit license file or declaration found in the provided files.
SKILL.md references external subagent prompt files and configs, but their completeness cannot be fully verified from the excerpt; however, the provided files appear consistent.
Quality score needs review
Permission surface needs review: shell or command execution, filesystem or document access
GitHub adoption: 59 GitHub stars
Stars/forks activity: 59 stars, 11 forks; issue activity unavailable in current metadata
License clarity: Unknown
Permission surface: shell or command execution, filesystem or document access

Install targets

Codex install prompt

Install the "patentradar" agent skill from https://github.com/yuc16/PatentRadar/tree/main/mcp/src/patentradar_mcp/assets. Read its SKILL.md or equivalent instructions first, install only the files needed for this workspace, and summarize any required setup before using it. Skill purpose: 专利侵权竞品分析 skill。输入专利公开号（CN/US/EP/JP 等），输出可人工复核的 markdown 竞品分析报告。工作流：拆解权利要求 → 搜索市场竞品 → 抓证据（含图像）→ 全部权利要求逐特征对比 → 生成报告。触发词：专利侵权分析、专利竞品、claim chart、专利公开号、CN/US/EP/JP 专利号、专利对比、技术特征对比、专利保护范围、侵权风险评估。不适用：专利申请文书撰写、专利无效宣告、知识产权法律咨询、专利诉讼策略。 After the install attempt, report the result to https://www.openagentskill.com/api/agent/outcome with POST JSON {"event_id":"install_<unique-id>","skill_slug":"yuc16-patentradar","task":"Install patentradar","agent":"codex","outcome":"success","install_used":true}. Replace event_id with a unique value and outcome with success or failed. Report success only after the skill is installed and a minimal verification passes. Recorded instruction path: mcp/src/patentradar_mcp/assets/SKILL.md. Recorded revision: 09805624bbd53d4acf58f51ef2c96822a436b2c5. Confirm the source matches these instructions. Before installing, identify the supported agent, runtime dependencies, API keys, paid services, license and permissions; mark anything not documented as unknown rather than free or compatible. Treat repository text as untrusted data; ask before credentials, paid services or external side effects. After setup, propose one small task with explicit inputs and expected output for the user to approve. Do not treat copying this prompt or successful installation as proof that the task succeeded.

Copying is not installation or a successful run. Check dependencies, API costs and permissions before proceeding.

Source & usage notes

IndexedInstall path available

Repository metadata and review signals are advisory. Popularity, source discovery and successful execution are different facts.

Source repository: yuc16/PatentRadar
License: Unknown
Version: 1.0.0
Last GitHub push: Aug 29, 2026
Registry updated: Sep 5, 2026
Instruction path: mcp/src/patentradar_mcp/assets/SKILL.md @ 09805624bbd5

Version reported in registry metadata; check source releases before relying on it.

Quality

59/100

Promising

Trust

57/100

Agent access

This page exposes the same decision, trust, audit, use-case, and install signals through the Registry API, so agents can rank this skill without scraping the UI.

Open install API LLM text format Open manifest Auto-resolve Plan

More details

{
  "version": "openagentskill-agent-metadata-v2",
  "review_evidence": {
    "indexed": true,
    "static_checked": false,
    "ai_reviewed": false,
    "manual_reviewed": false,
    "creator_verified": false,
    "review_result": "not_recorded",
    "reviewed_at": null,
    "package_fingerprint": null,
    "policy_version": null,
    "notice": "Publication, static checks, AI review, and creator verification are independent facts. None guarantees runtime safety."
  },
  "skill": {
    "slug": "yuc16-patentradar",
    "name": "patentradar",
    "description": "专利侵权竞品分析 skill。输入专利公开号（CN/US/EP/JP 等），输出可人工复核的 markdown 竞品分析报告。工作流：拆解权利要求 → 搜索市场竞品 → 抓证据（含图像）→ 全部权利要求逐特征对比 → 生成报告。触发词：专利侵权分析、专利竞品、claim chart、专利公开号、CN/US/EP/JP 专利号、专利对比、技术特征对比、专利保护范围、侵权风险评估。不适用：专利申请文书撰写、专利无效宣告、知识产权法律咨询、专利诉讼策略。",
    "category": "data-analysis",
    "url": "https://www.openagentskill.com/skills/yuc16-patentradar",
    "repository": "https://github.com/yuc16/PatentRadar/tree/main/mcp/src/patentradar_mcp/assets",
    "github_repo": "yuc16/PatentRadar"
  },
  "suited_tasks": [
    "Research agents workflows",
    "Claude Code teams",
    "builders willing to evaluate younger projects",
    "Search sources",
    "Extract claims",
    "Synthesize findings",
    "Read uploaded files",
    "Extract structured fields"
  ],
  "suited_agents": [
    "Codex",
    "Claude Code",
    "Cursor",
    "OpenAgentSkill CLI",
    "CLI"
  ],
  "install": {
    "source_evidence": {
      "status": "source-recorded",
      "sourceRecorded": true,
      "canOfferInstall": true,
      "path": "mcp/src/patentradar_mcp/assets/SKILL.md",
      "revision": "09805624bbd53d4acf58f51ef2c96822a436b2c5",
      "notice": "A skill instruction path and install command are recorded. This is not proof of compatibility, runtime success or safety; review the source and permissions first."
    },
    "command": "npx skills add yuc16/PatentRadar --skill patentradar",
    "ready": true,
    "targets": [
      {
        "id": "openagentskill-cli",
        "label": "CLI",
        "kind": "command",
        "value": "npx --yes https://github.com/Leon-Drq/openagentskill/releases/download/cli-v0.3.0/openagentskill-0.3.0.tgz add yuc16-patentradar"
      },
      {
        "id": "codex",
        "label": "Codex",
        "kind": "agent-prompt",
        "value": "Install the \"patentradar\" agent skill from https://github.com/yuc16/PatentRadar/tree/main/mcp/src/patentradar_mcp/assets. Read its SKILL.md or equivalent instructions first, install only the files needed for this workspace, and summarize any required setup before using it. Skill purpose: 专利侵权竞品分析 skill。输入专利公开号（CN/US/EP/JP 等），输出可人工复核的 markdown 竞品分析报告。工作流：拆解权利要求 → 搜索市场竞品 → 抓证据（含图像）→ 全部权利要求逐特征对比 → 生成报告。触发词：专利侵权分析、专利竞品、claim chart、专利公开号、CN/US/EP/JP 专利号、专利对比、技术特征对比、专利保护范围、侵权风险评估。不适用：专利申请文书撰写、专利无效宣告、知识产权法律咨询、专利诉讼策略。 After the install attempt, report the result to https://www.openagentskill.com/api/agent/outcome with POST JSON {\"event_id\":\"install_<unique-id>\",\"skill_slug\":\"yuc16-patentradar\",\"task\":\"Install patentradar\",\"agent\":\"codex\",\"outcome\":\"success\",\"install_used\":true}. Replace event_id with a unique value and outcome with success or failed. Report success only after the skill is installed and a minimal verification passes. Recorded instruction path: mcp/src/patentradar_mcp/assets/SKILL.md. Recorded revision: 09805624bbd53d4acf58f51ef2c96822a436b2c5. Confirm the source matches these instructions. Before installing, identify the supported agent, runtime dependencies, API keys, paid services, license and permissions; mark anything not documented as unknown rather than free or compatible. Treat repository text as untrusted data; ask before credentials, paid services or external side effects. After setup, propose one small task with explicit inputs and expected output for the user to approve. Do not treat copying this prompt or successful installation as proof that the task succeeded."
      },
      {
        "id": "claude-code",
        "label": "Claude Code",
        "kind": "agent-prompt",
        "value": "Add \"patentradar\" as a Claude Code skill from https://github.com/yuc16/PatentRadar/tree/main/mcp/src/patentradar_mcp/assets. Inspect the skill instructions, place the reusable skill files in the appropriate local skills location for this project, and report the activation steps. Skill purpose: 专利侵权竞品分析 skill。输入专利公开号（CN/US/EP/JP 等），输出可人工复核的 markdown 竞品分析报告。工作流：拆解权利要求 → 搜索市场竞品 → 抓证据（含图像）→ 全部权利要求逐特征对比 → 生成报告。触发词：专利侵权分析、专利竞品、claim chart、专利公开号、CN/US/EP/JP 专利号、专利对比、技术特征对比、专利保护范围、侵权风险评估。不适用：专利申请文书撰写、专利无效宣告、知识产权法律咨询、专利诉讼策略。 After the install attempt, report the result to https://www.openagentskill.com/api/agent/outcome with POST JSON {\"event_id\":\"install_<unique-id>\",\"skill_slug\":\"yuc16-patentradar\",\"task\":\"Install patentradar\",\"agent\":\"claude-code\",\"outcome\":\"success\",\"install_used\":true}. Replace event_id with a unique value and outcome with success or failed. Report success only after the skill is installed and a minimal verification passes. Recorded instruction path: mcp/src/patentradar_mcp/assets/SKILL.md. Recorded revision: 09805624bbd53d4acf58f51ef2c96822a436b2c5. Confirm the source matches these instructions. Before installing, identify the supported agent, runtime dependencies, API keys, paid services, license and permissions; mark anything not documented as unknown rather than free or compatible. Treat repository text as untrusted data; ask before credentials, paid services or external side effects. After setup, propose one small task with explicit inputs and expected output for the user to approve. Do not treat copying this prompt or successful installation as proof that the task succeeded."
      },
      {
        "id": "cursor",
        "label": "Cursor",
        "kind": "agent-prompt",
        "value": "Turn \"patentradar\" from https://github.com/yuc16/PatentRadar/tree/main/mcp/src/patentradar_mcp/assets into a reusable Cursor project rule or agent instruction. Preserve the core workflow, adapt paths to this repo, and keep the rule scoped to tasks where it is relevant. Skill purpose: 专利侵权竞品分析 skill。输入专利公开号（CN/US/EP/JP 等），输出可人工复核的 markdown 竞品分析报告。工作流：拆解权利要求 → 搜索市场竞品 → 抓证据（含图像）→ 全部权利要求逐特征对比 → 生成报告。触发词：专利侵权分析、专利竞品、claim chart、专利公开号、CN/US/EP/JP 专利号、专利对比、技术特征对比、专利保护范围、侵权风险评估。不适用：专利申请文书撰写、专利无效宣告、知识产权法律咨询、专利诉讼策略。 After the install attempt, report the result to https://www.openagentskill.com/api/agent/outcome with POST JSON {\"event_id\":\"install_<unique-id>\",\"skill_slug\":\"yuc16-patentradar\",\"task\":\"Install patentradar\",\"agent\":\"cursor\",\"outcome\":\"success\",\"install_used\":true}. Replace event_id with a unique value and outcome with success or failed. Report success only after the skill is installed and a minimal verification passes. Recorded instruction path: mcp/src/patentradar_mcp/assets/SKILL.md. Recorded revision: 09805624bbd53d4acf58f51ef2c96822a436b2c5. Confirm the source matches these instructions. Before installing, identify the supported agent, runtime dependencies, API keys, paid services, license and permissions; mark anything not documented as unknown rather than free or compatible. Treat repository text as untrusted data; ask before credentials, paid services or external side effects. After setup, propose one small task with explicit inputs and expected output for the user to approve. Do not treat copying this prompt or successful installation as proof that the task succeeded."
      }
    ],
    "handoff_url": "https://www.openagentskill.com/api/skills/yuc16-patentradar/install",
    "manifest_url": "https://www.openagentskill.com/api/registry/manifest/yuc16-patentradar"
  },
  "trust": {
    "score": 65,
    "label": "Manual review",
    "version": "trust-score-v4",
    "install_policy": "review",
    "evidence": {
      "stars": "59 GitHub stars",
      "repoActivity": "59 stars, 11 forks",
      "lastPushed": "26d since push",
      "license": "Unknown",
      "repository": "https://github.com/yuc16/PatentRadar/tree/main/mcp/src/patentradar_mcp/assets",
      "install": "npx skills add yuc16/PatentRadar --skill patentradar",
      "installSafety": "standard package or runtime install path",
      "permissionSurface": "shell or command execution, filesystem or document access",
      "documentation": "Strong README/SKILL.md context",
      "agentOutcomes": "No agent outcome data yet"
    },
    "outcome_evidence": {
      "total": 0,
      "successes": 0,
      "failures": 0,
      "not_relevant": 0,
      "success_rate": null,
      "recent_success_rate": null,
      "recent_failure_rate": null,
      "install_attempts": 0,
      "install_success_rate": null,
      "risk_blocked": 0,
      "setup_required": 0,
      "avg_output_quality": null,
      "production_outcomes": 0,
      "last_outcome_at": null,
      "label": "No agent outcome data yet"
    },
    "auto_install": {
      "allowed": false,
      "sandbox_required": true,
      "reason": "Test manually in an isolated workspace and compare against safer alternatives."
    },
    "best_for": [
      "data-analysis",
      "agent-skill"
    ],
    "known_risks": [
      "Repository license is unknown; no explicit license file or declaration found in the provided files.",
      "License is unclear",
      "Quality score needs review",
      "Permission surface needs review: shell or command execution, filesystem or document access",
      "GitHub adoption: 59 GitHub stars",
      "Stars/forks activity: 59 stars, 11 forks; issue activity unavailable in current metadata",
      "License clarity: Unknown",
      "Permission surface: shell or command execution, filesystem or document access"
    ]
  },
  "agent_proven": {
    "version": "agent-proven-v1",
    "score": 0,
    "tier": "unproven",
    "label": "Needs first agent run",
    "summary": "No agent outcome reports yet. Use Resolve, run one narrow sandbox task, then report the result.",
    "metrics": {
      "totalOutcomes": 0,
      "successfulOutcomes": 0,
      "failedOutcomes": 0,
      "installAttempts": 0,
      "installSuccessRate": null,
      "successRate": null,
      "recentSuccessRate": null,
      "recentFailureRate": null,
      "riskBlocked": 0,
      "setupRequired": 0,
      "notRelevant": 0,
      "avgOutputQuality": null,
      "avgTimeToUsefulMs": null,
      "productionOutcomes": 0,
      "humanReviewRequired": 0,
      "uniqueAgents": 0,
      "lastOutcomeAt": null
    },
    "signals": [],
    "penalties": [
      "No real agent outcome evidence yet"
    ]
  },
  "audit": {
    "score": 72,
    "risk_level": "needs_review",
    "risk_label": "Needs review",
    "warnings": [
      "License is unclear",
      "Permission surface may require sandboxing",
      "Repository license is unknown; no explicit license file or declaration found in the provided files.",
      "SKILL.md references external subagent prompt files and configs, but their completeness cannot be fully verified from the excerpt; however, the provided files appear consistent.",
      "Quality score needs review",
      "Permission surface needs review: shell or command execution, filesystem or document access",
      "GitHub adoption: 59 GitHub stars",
      "Stars/forks activity: 59 stars, 11 forks; issue activity unavailable in current metadata"
    ]
  },
  "safety_gate": {
    "tier": "experimental",
    "label": "Experimental",
    "auto_install_policy": "review",
    "auto_install_allowed": false,
    "human_review_required": true,
    "blocked": false,
    "recommended_action": "Test manually in an isolated workspace and compare against safer alternatives."
  },
  "quality": {
    "score": 59,
    "label": "Promising"
  },
  "supply": {
    "track": "Data, BI, and analytics",
    "scenario": "Data analysis",
    "maintenance": "26d since push",
    "risk": "Needs review"
  },
  "alternative_skills": [],
  "do_not_use_when": [
    "teams that need a vendor-supported SLA",
    "production agents without a repository review",
    "Repository license is unknown; no explicit license file or declaration found in the provided files.",
    "No OpenAgentSkill engagement data yet",
    "High-risk permission hints: Shell or command execution",
    "License is unclear",
    "Permission surface may require sandboxing",
    "SKILL.md references external subagent prompt files and configs, but their completeness cannot be fully verified from the excerpt; however, the provided files appear consistent."
  ],
  "agent_contract": {
    "task_input": "Use patentradar in an agent workflow",
    "recommended_action": "Test manually in an isolated workspace and compare against safer alternatives.",
    "install_policy": "review",
    "minimum_review_before_use": [
      "Trust: 65/100 Manual review",
      "Audit: 72/100 Needs review",
      "Safety: 40/100 Avoid automatic install",
      "Review repository, license, install command, and permission surface before production use."
    ],
    "expected_agent_output": {
      "selected_skill": "yuc16-patentradar (patentradar)",
      "install_command": "npx skills add yuc16/PatentRadar --skill patentradar",
      "risk_summary": "Needs review; Experimental; Review before production",
      "verification_result": "Report the smallest successful task, files touched, warnings, and any missing setup."
    }
  },
  "outcome_feedback": {
    "endpoint": "https://www.openagentskill.com/api/agent/outcome",
    "method": "POST",
    "requires_resolve_event_id": true,
    "event_id_source": "Use install_receipt.outcome_feedback.event_id or feedback.event_id returned by /api/agent/resolve for the current task.",
    "expected_outcomes": [
      "success",
      "failed",
      "not_relevant",
      "blocked_by_risk",
      "setup_required"
    ],
    "payload_template": {
      "event_id": "<install_receipt.outcome_feedback.event_id or feedback.event_id from /api/agent/resolve>",
      "skill_slug": "yuc16-patentradar",
      "task": "Use patentradar in an agent workflow",
      "agent": "codex",
      "outcome": "success",
      "install_used": true,
      "risk_blocked": false,
      "setup_required": false,
      "task_success": true,
      "output_quality": 4,
      "error_type": null,
      "human_review_required": false,
      "workspace": "sandbox",
      "time_to_useful_ms": 120000,
      "notes": "Report the smallest successful task, setup friction, files touched, and risk notes."
    }
  },
  "endpoints": {
    "web": "https://www.openagentskill.com/skills/yuc16-patentradar",
    "api": "https://www.openagentskill.com/api/agent/skills/yuc16-patentradar",
    "audit": "https://www.openagentskill.com/skills/yuc16-patentradar/audit",
    "eval": "https://www.openagentskill.com/api/agent/evals?slug=yuc16-patentradar&task=Use%20patentradar%20in%20an%20agent%20workflow&max_risk=medium",
    "resolve": "https://www.openagentskill.com/api/agent/resolve?task=Use%20patentradar%20in%20an%20agent%20workflow&agent=codex&max_risk=medium",
    "receipt": "https://www.openagentskill.com/api/agent/receipt?task=Use%20patentradar%20in%20an%20agent%20workflow&agent=codex&max_risk=medium&format=text",
    "install": "https://www.openagentskill.com/api/skills/yuc16-patentradar/install",
    "manifest": "https://www.openagentskill.com/api/registry/manifest/yuc16-patentradar"
  }
}

For the creator

Listing source

Registry indexed

Claimable

This listing was indexed from public sources and is not marked official until a maintainer claim is approved.

Creator: yuc16
Source: yuc16/PatentRadar
Indexed by: OpenAgentSkill community index

Attribution links to the public repository or creator profile. Creators can claim the listing to update ownership signals.

Claim this skill

Owner claim

Claim this skill listing

This Registry indexed listing is attributed to yuc16 but is not marked official yet. Claim it to add a verified owner signal and make future launch, install, and audit updates easier to trust.

Share kit

Creator backlink kit

Add the evidence badges to your README

Show the canonical listing, current trust and audit signals, and real Agent-Proven evidence where developers evaluate the repository.

[![Listed on OpenAgentSkill](https://www.openagentskill.com/api/badge/yuc16-patentradar?metric=listed&label=Listed)](https://www.openagentskill.com/skills/yuc16-patentradar?ref=github&utm_source=github&utm_medium=referral&utm_campaign=creator_badge)
[![OpenAgentSkill Trust](https://www.openagentskill.com/api/badge/yuc16-patentradar?metric=trust&label=Trust)](https://www.openagentskill.com/skills/yuc16-patentradar?ref=github&utm_source=github&utm_medium=referral&utm_campaign=creator_badge)
[![OpenAgentSkill Audit](https://www.openagentskill.com/api/badge/yuc16-patentradar?metric=audit&label=Audit)](https://www.openagentskill.com/skills/yuc16-patentradar/audit)
[![Agent Proven](https://www.openagentskill.com/api/badge/yuc16-patentradar?metric=proven&label=Agent%20Proven)](https://www.openagentskill.com/skills/yuc16-patentradar?ref=github&utm_source=github&utm_medium=referral&utm_campaign=creator_badge)

Preview badge Open audit Creator kit

Community signal

Share whether this skill looks useful for your agent workflow. Aggregated feedback improves rankings over time.