agentsclimarketplace

Yao geo brand graph

Skill ShenyuanNext/wepr-growth-skills/skills/yao-geo-brand-graph

当用户需要把企业资料、官网、产品、客户案例、人物、地点、证据源和场景词整理成品牌实体知识图谱时使用;用于知识库升级、页面结构设计、监测纠偏、品牌百科化和 AI 内容一致性治理。From its SKILL.md

Install
npx -y skills add ShenyuanNext/wepr-growth-skills --skill yao-geo-brand-graph

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

2 things to look at

  • 28 days oldThe repository was created 28 days ago. New is not bad, but a brand new repository carrying a familiar-sounding name is the shape a typosquat arrives in, and there has been no time for anyone else to find a problem with it.
  • 2 stars2 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

SKILL.md

2.8 KB, 876 tokens by cl100k_base, as published. Nobody here has run it

<!-- Copyright © 2026 姚金刚. All rights reserved. Project: yao-geo-brand-graph Created by: 姚金刚 Date: 2026-05-16 X: https://x.com/yaojingang -->

Yao GEO Brand Graph

执行流程

  1. references/skill-method.md 建立来源账本,区分官网、官方文档、投资者关系、客户案例、第三方资料和待确认素材。
  2. references/real-data-acquisition.md 对来源 URL 做自动连通核验;能访问的来源进入真实数据样本,受限来源进入人工复核。
  3. 按 mention -> candidate -> canonical entity 做实体消歧。
  4. references/entity-schema.md 抽取品牌、产品、服务、功能、技术、行业、用户、场景、客户、案例、证据、地点、时间。
  5. references/evidence-policy.md 抽取有方向的关系边,每条关系必须有证据 ID。
  6. 先做完整性自检:权威参考、来源覆盖、真实数据来源核验、实体覆盖、关系证据、Schema 一致性、内容资产对齐、国内 AI 平台监测闭环和隐私授权缺口。
  7. 输出实体清单、关系清单、可信等级、消歧表、Mermaid、JSON-LD、RDF 三元组、国内 AI 平台测试场景和补强建议。

排版质量门

  • Word 必须使用横向页面、真实 w:tbl 表格、固定 dxa 表宽,禁止自动表宽撑出页面。
  • Word 表格网格总宽必须小于正文宽度,并保留右侧安全边距。
  • URL、英文实体 ID、英文产品名和长英文短语必须在单元格内主动断行。
  • HTML/PDF 按 kami paper long-doc 风格排版:暖米纸底、ivory 内容面、油墨蓝点缀、暖灰边框、serif 标题、紧凑长文档节奏。
  • HTML 可视化报告必须有固定跟随的目录菜单;PDF 打印版可隐藏菜单以保持正文排版。
  • 报告必须包含系统化分析模块,不能只输出基础实体表和图谱图。

工具入口

  • scripts/render_yao_geo_brand_graph.py:从结构化 report_input.json 生成 Word、PDF、HTML、Markdown 和 quality-report.json
  • scripts/collect_source_validation.py:读取来源账本,自动采样 URL 可达性、页面标题和核验时间,并可写回 source_validation
  • examples/hubspot-domestic-ai-test/report_input.json:HubSpot 国内 AI 平台测试样例输入。
  • templates/brief-template.md:项目输入简报模板。
  • evals/expected_artifacts.json:必要文件和示例交付检查清单。
  • reports/output-risk-profile.mdreports/artifact-design-profile.md:输出风险和排版设计约束。

Keep looking

Skills are one crate of 326,852. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.