SuperAPI AI Directory
Partner with us中文AI newsFavorites
Skillsyao-geo-execution-roadmap

yao-geo-execution-roadmap

General & other

View repository

Updated 10-16 09:52

About this skill

Yao GEO Execution Roadmap

Use cases

  • There is already a GEO diagnosis, opportunity map, platform sample, or brand fact base, and it needs to be turned into a 30/60/90-day execution plan.
  • The user needs to give the CEO, marketing lead, growth team, content team, technical team, or client delivery team a clear view of tasks, budget, dependencies, and acceptance criteria.
  • The project needs to cover the official website pages, content matrix, title system, knowledge base, external evidence, and monitoring loop at the same time.

Inapplicable scenarios

  • When there is no basic information about the brand, questions, platforms, pages, or competitors, do not directly generate an implementation plan; first supplement the brief or switch to panoramic diagnosis.
  • When only an AI search question set, intent clusters, title library, or a single article is needed, the corresponding content/intent skill should be used.
  • When only effect tracking fields, dashboards, or attribution design are needed, use yao-geo-tracking。

Required inputs

  • Panoramic diagnosis report, opportunity map, or equivalent diagnosis summary.
  • Target platform scope: by default covers DeepSeek, Doubao, Qwen, Kimi, and Yuanbao.
  • Business objectives: leads, deals, exposure, brand authority, investment promotion, product education, or customer success.
  • Resource constraints: content capacity, development schedule, budget, data permissions, legal review, and delivery team boundaries.
  • Time range: by default 30 days for quick wins, 60 days for asset building, and 90 days for closed-loop optimization.

Required reading materials

  • references/skill-method.md
  • references/real-data-capability.md
  • references/analysis-completeness-framework.md
  • references/research-backed-framework.md
  • references/cn-platform-adaptation.md
  • references/artifact-layout.md
  • references/quality-gates.md

Execution process

  1. Build on the diagnosis, do not start from a blank checklist: first extract brand facts, question clusters, platform gaps, page and content gaps, evidence source gaps, and monitoring metric gaps.
  2. Determine the real data mode: user-provided, public web verification, platform sample import, or hybrid mode; clearly state capability boundaries, source freshness, and gaps still needing evidence.
  3. Output an analysis completeness matrix: the ten dimensions of business objectives, question coverage, brand entity, page technology, content assets, title system, knowledge base, external evidence, platform adaptation, and monitoring governance all need current status, gaps, actions, acceptance criteria, and responsible roles.
  4. Output evidence maturity and fact base: classify core assertions into A officially verified, B third-party corroborable, C internal pending authorization, D needs additional evidence, and E prohibited for use, and provide handling actions for D/E.
  5. Determine the North Star metrics: select at least 3 from brand recommendation rate, citation rate, answer accuracy rate, core question coverage rate, evidence support rate, and lead contribution.
  6. Break down into six project packages: page technology, content matrix, title system, knowledge base, external evidence, and monitoring loop.
  7. Each project package must clearly state objectives, tasks, inputs, steps, owners, deliverables, acceptance metrics, dependencies, priorities, and resource budget.
  8. Output a 30/60/90-day roadmap: first do verifiable quick wins, then do knowledge assets and batch content, and finally do monitoring, review, and iteration, and provide resource trade-offs.
  9. Differentiated adaptation for domestic platforms: DeepSeek focuses on logic and evidence chains, Doubao focuses on colloquial questions and creative answers, Qwen focuses on traceable citations and retrieval configuration, Kimi focuses on long texts and deep research, and Yuanbao focuses on the WeChat ecosystem and Tencent content entry points.
  10. Add risk contingency plans, governance cadence, and decision mechanisms; generate reports in four formats and self-check. HTML has a fixed menu; Word has fixed table geometry, with long tables split into narrow tables.

Output Contract

  • GEO 综合实施方案
  • 真实数据能力与采集计划
  • 分析完整性矩阵
  • 证据成熟度与事实底座
  • 90 天执行路线图
  • 角色分工与验收指标表
  • 资源预算与优先级
  • 治理节奏与决策机制
  • 内容矩阵、页面改造、知识库建设和监测计划
  • 国内 AI 平台差异化执行动作
  • 风险预案与合规节点
  • Default four-piece set: Word (.docx）、PDF（.pdf), HTML with a fixed follow-along menu bar (.html）、Markdown（.md）

Method Constraints

  • The plan must build on diagnostic data and must not remain a generic checklist.
  • All tasks must have acceptance metrics and responsible roles.
  • All reports must first provide an analysis completeness matrix, then provide the roadmap and tasks.
  • All real data must clearly state the source type, evidence level, verification date, and freshness; data that cannot be obtained goes into the evidence supplementation plan.
  • All budgets, schedules, and resource investments must clearly state priorities, trade-offs, and escalation conditions.
  • You cannot promise that a certain platform will definitely cite a certain page; you can only improve discoverability, verifiability, and extractability.
  • All brand facts, customer data, product capabilities, prices, cases, qualifications, and financial data must have sources; if they cannot be confirmed, mark them as pending confirmation.
  • Default output is Simplified Chinese; retain brand names, product names, URLs, platform names, and English proper nouns.

Generator

Use the built-in script to generate reports in four formats:

python3 scripts/render_yao_geo_execution_roadmap.py \
  --input examples/hubspot-cn-demo/report_input.json \
  --output-dir examples/hubspot-cn-demo

After generation, read quality-report.json, all key checks must be pass: true。