Agent skill

Geo

by liangdabiao in liangdabiao/GEO-Content-Optimizer-Skill

完整的 GEO（生成式引擎优化）服务流水线：给一个产品官网 URL 和介绍材料，做站点诊断与 AI 答案采样、生成带验收标准的执行工单、产出可直接部署的资产（llms.txt / JSON-LD / 定义块 / FAQ / 内容大纲与初稿）、自动验收工单是否闭环、并打包成可直接发给客户的交付物。可按周期复跑，做长期 GEO 运营与月报。

MITAuto-check: notesMarketing & SEO

Install Geo

skills CLI

$ npx skills add liangdabiao/GEO-Content-Optimizer-Skill --skill geo -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI

$ gh skill install liangdabiao/GEO-Content-Optimizer-Skill geo --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy

$ git clone --depth 1 https://github.com/liangdabiao/GEO-Content-Optimizer-Skill.git skills-src && mkdir -p .claude/skills && cp -r skills-src/skills/geolook .claude/skills/geo && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Paste into Claude Code

Install the "geo" agent skill from https://github.com/liangdabiao/GEO-Content-Optimizer-Skill/tree/main/skills/geolook into .claude/skills/geo/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "geo", then confirm the skill loads.

Claude Code copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.

$skill-installer

$skill-installer install https://github.com/liangdabiao/GEO-Content-Optimizer-Skill/tree/main/skills/geolook

Type this inside Codex. $skill-installer <name> installs a curated skill from openai/skills. The installer writes to $CODEX_HOME/skills (default ~/.codex/skills). Restart Codex if the skill does not show up.

skills CLI

$ npx skills add liangdabiao/GEO-Content-Optimizer-Skill --skill geo -a codex

Project install goes to .agents/skills/; add -g for ~/.codex/skills/.

GitHub CLI

$ gh skill install liangdabiao/GEO-Content-Optimizer-Skill geo --agent codex

Project scope by default (.agents/skills/); add --scope user for a personal install.

Manual copy

$ git clone --depth 1 https://github.com/liangdabiao/GEO-Content-Optimizer-Skill.git skills-src && mkdir -p .agents/skills && cp -r skills-src/skills/geolook .agents/skills/geo && rm -rf skills-src

Use ~/.agents/skills/ instead of .agents/skills for a personal install.

Codex skills documentation · loads skills from .agents/skills/

Paste into Codex

Install the "geo" agent skill from https://github.com/liangdabiao/GEO-Content-Optimizer-Skill/tree/main/skills/geolook into .agents/skills/geo/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "geo", then confirm the skill loads.

Codex copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.

skills CLI

$ npx skills add liangdabiao/GEO-Content-Optimizer-Skill --skill geo -a cursor

Project install goes to .agents/skills/; add -g for ~/.cursor/skills/.

GitHub CLI

$ gh skill install liangdabiao/GEO-Content-Optimizer-Skill geo --agent cursor

Project scope by default (.agents/skills/); add --scope user for a personal install.

Manual copy

$ git clone --depth 1 https://github.com/liangdabiao/GEO-Content-Optimizer-Skill.git skills-src && mkdir -p .cursor/skills && cp -r skills-src/skills/geolook .cursor/skills/geo && rm -rf skills-src

Use ~/.cursor/skills/ instead of .cursor/skills for a personal install.

Cursor skills documentation · loads skills from .cursor/skills/, .agents/skills/, .claude/skills/, .codex/skills/

Paste into Cursor

Install the "geo" agent skill from https://github.com/liangdabiao/GEO-Content-Optimizer-Skill/tree/main/skills/geolook into .cursor/skills/geo/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "geo", then confirm the skill loads.

Cursor copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.

gemini skills install

$ gemini skills install https://github.com/liangdabiao/GEO-Content-Optimizer-Skill.git --path skills/geolook

--scope user (default) or --scope workspace; --path is the subfolder of the repo that holds the skill; --consent skips the security confirmation prompt.

skills CLI

$ npx skills add liangdabiao/GEO-Content-Optimizer-Skill --skill geo -a gemini-cli

Project install goes to .agents/skills/; add -g for ~/.gemini/skills/.

GitHub CLI

$ gh skill install liangdabiao/GEO-Content-Optimizer-Skill geo --agent gemini-cli

Project scope by default (.agents/skills/); add --scope user for a personal install.

Manual copy

$ git clone --depth 1 https://github.com/liangdabiao/GEO-Content-Optimizer-Skill.git skills-src && mkdir -p .gemini/skills && cp -r skills-src/skills/geolook .gemini/skills/geo && rm -rf skills-src

Use ~/.gemini/skills/ instead of .gemini/skills for a personal install, then run /skills reload.

Gemini CLI skills documentation · loads skills from .gemini/skills/, .agents/skills/

Paste into Gemini CLI

Install the "geo" agent skill from https://github.com/liangdabiao/GEO-Content-Optimizer-Skill/tree/main/skills/geolook into .gemini/skills/geo/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "geo", then confirm the skill loads.

Gemini CLI copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.

GitHub CLI

$ gh skill install liangdabiao/GEO-Content-Optimizer-Skill geo

Installs for Copilot at project scope by default; add --scope user for a personal install. Preview a skill first with gh skill preview. Needs GitHub CLI 2.90.0 or later (public preview).

skills CLI

$ npx skills add liangdabiao/GEO-Content-Optimizer-Skill --skill geo -a github-copilot

Project install goes to .agents/skills/; add -g for ~/.copilot/skills/.

Manual copy

$ git clone --depth 1 https://github.com/liangdabiao/GEO-Content-Optimizer-Skill.git skills-src && mkdir -p .github/skills && cp -r skills-src/skills/geolook .github/skills/geo && rm -rf skills-src

Use ~/.copilot/skills/ instead of .github/skills for a personal install. Commit .github/skills so cloud agent and code review can use it.

GitHub Copilot skills documentation · loads skills from .github/skills/, .claude/skills/, .agents/skills/

Paste into GitHub Copilot

Install the "geo" agent skill from https://github.com/liangdabiao/GEO-Content-Optimizer-Skill/tree/main/skills/geolook into .github/skills/geo/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "geo", then confirm the skill loads.

GitHub Copilot copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.

skills CLI

$ npx skills add liangdabiao/GEO-Content-Optimizer-Skill --skill geo -a opencode

OpenCode documents no install command of its own. Project install goes to .agents/skills/; add -g for ~/.config/opencode/skills/.

GitHub CLI

$ gh skill install liangdabiao/GEO-Content-Optimizer-Skill geo --agent opencode

Project scope by default (.agents/skills/); add --scope user for a personal install.

Manual copy

$ git clone --depth 1 https://github.com/liangdabiao/GEO-Content-Optimizer-Skill.git skills-src && mkdir -p .opencode/skills && cp -r skills-src/skills/geolook .opencode/skills/geo && rm -rf skills-src

Use ~/.config/opencode/skills/ instead of .opencode/skills for a personal install.

OpenCode skills documentation · loads skills from .opencode/skills/, .claude/skills/, .agents/skills/

Paste into OpenCode

Install the "geo" agent skill from https://github.com/liangdabiao/GEO-Content-Optimizer-Skill/tree/main/skills/geolook into .opencode/skills/geo/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "geo", then confirm the skill loads.

OpenCode copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.

Facts

Skill name: geo
GitHub stars: 205
Used in: 1 other repo
Token cost: ~2.3k tokens
SKILL.md length: 497 words
Files: 63 (incl. scripts, references)
Skills in repo: 3
Repo updated: Aug 5, 2026
First seen: Oct 7, 2026
Licence: MIT

At a glance

Works in 4 steps: SPA 空壳页（word_count 接近 0）——国内官网最常见致命伤，AI… → robots 有没有封 AI 抓取器——封了就什么都别谈 → language_coverage 中英是否对等——做海外却没英文原生页直接 P0 → …
Tasks that involve AI search optimization
SKILL.md covers 最快的用法：给一个网址, 两种用法, 先读什么 and 步骤 0 · 建项目, plus 7 more sections
Calls python3; needs ZHIPUAI_API_KEY and ARK_API_KEY

What it does

Geo is an agent skill from liangdabiao/GEO-Content-Optimizer-Skill. 完整的 GEO（生成式引擎优化）服务流水线：给一个产品官网 URL 和介绍材料，做站点诊断与 AI 答案采样、生成带验收标准的执行工单、产出可直接部署的资产（llms.txt / JSON-LD / 定义块 / FAQ / 内容大纲与初稿）、自动验收工单是否闭环、并打包成可直接发给客户的交付物。可按周期复跑，做长期 GEO 运营与月报。国内和海外双市场并行：国内覆盖智谱GLM/豆包/DeepSeek/Kimi/MiniMax/纳米AI/百度AI，海外覆盖 Gemini/ChatGPT/Claude/Grok/Perplexity，问题库与指标按市场分开算。当用户说「做 GEO」「生成式引擎优化」「让 AI 推荐我的产品」「AI 搜索里搜不到我们」「GEO 方案/诊断/监测/月报/交付」「给客户做 GEO 服务」，或给一个产品网址要做 AI 可见性优化时使用。GEO 指生成式引擎优化，不是地理信息；不用于传统 SEO 关键词排名、竞价投放或建站。

Its SKILL.md is about 2.3k tokens, which your agent loads only when the skill is triggered. The skill folder holds 64 other files, including scripts and reference files (for example `README.md`, `docs/302ai-integration-research.md` and `docs/ARCHITECTURE.md`).

It sits in Marketing & SEO, covering AI search optimization, Web search and Schema markup. It works with DeepSeek, Kimi, MiniMax and OpenAI. The repository describes itself as: GEO（Generative Engine Optimization）是面向 AI 搜索引擎的内容优化方法论。就像 SEO 优化 Google 排名，GEO 优化你的内容在 ChatGPT、Perplexity、Gemini、Google AI Overview 等 AI 引擎中的引用率。本项目提供3个 Agent Skill，覆盖 GEO 全流程：. The licence is MIT.

When your agent uses it

Tasks that involve AI search optimization
Tasks that involve Web search
Tasks that involve Schema markup

Example prompts

“/geo”

Requirements

Python 3
A credential in ZHIPUAI_API_KEY
A credential in ARK_API_KEY

Workflow steps

4 steps, taken from the first numbered list in SKILL.md.

SPA 空壳页（word_count 接近 0）——国内官网最常见致命伤，AI 抓取器看到的是空白
robots 有没有封 AI 抓取器——封了就什么都别谈
language_coverage 中英是否对等——做海外却没英文原生页直接 P0
block_gap 缺得最多的块——内容工程第一优先级

What it can do on your machine

Read from SKILL.md and the folder at commit f56124f. It shows what the files ask for, not the result of running them.

Tool permissions
Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.
From allowed-tools in the SKILL.md frontmatter.
Runs code
Ships 1 file in scripts/, which the agent can run.
Shell commands in SKILL.md call:
- python3
From the folder's file list and the shell code blocks in SKILL.md.
Network
No URLs in SKILL.md.
From URLs in SKILL.md, links to its own repository left out.
Credentials
Names these keys or tokens, usually read from environment variables:
- ZHIPUAI_API_KEY
- ARK_API_KEY
- DEEPSEEK_API_KEY
- MOONSHOT_API_KEY
- MINIMAX_API_KEY
- GEMINI_API_KEY
- OPENAI_API_KEY
- ANTHROPIC_API_KEY
- XAI_API_KEY
- PERPLEXITY_API_KEY
From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Geo loads about 2.3k tokens when it runs, and up to ~13k if it reads all its reference files. Until then it costs about 110 tokens; SKILL.md has 497 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it: ~110
When it runs · the whole SKILL.md, loaded when a task matches: ~2.3k
With references · SKILL.md plus every file in references/, read only if the agent opens them: ~13k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check: notes

The automated check noted patterns worth knowing about, such as sudo or a known installer.

NoteMentions a .env fileSKILL.md:169

Key 放项目根目录 `.env`（已 gitignore，权限 600），脚本自动加载。

NoteMentions a .env fileSKILL.md:171

豆包在 302.AI 模式下走「搜索+回答」组合，单题超时读 `.env` 的 `AI302AI_SEARCH_ASK_TIMEOUT`

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); the scripts in this folder are not scanned.

SKILL.md

The full file from liangdabiao/GEO-Content-Optimizer-Skill at commit f56124f, republished under its MIT licence (© liangdabiao). 497 words, ~2,296 tokens.

Download SKILL.mdSave it as .claude/skills/geo/SKILL.md (or your agent's skills folder). This skill also uses 62 other files; get the full folder from GitHub.

name: geo
description: 完整的 GEO（生成式引擎优化）服务流水线：给一个产品官网 URL 和介绍材料，做站点诊断与 AI 答案采样、生成带验收标准的执行工单、产出可直接部署的资产（llms.txt / JSON-LD / 定义块 / FAQ / 内容大纲与初稿）、自动验收工单是否闭环、并打包成可直接发给客户的交付物。可按周期复跑，做长期 GEO 运营与月报。国内和海外双市场并行：国内覆盖智谱GLM/豆包/DeepSeek/Kimi/MiniMax/纳米AI/百度AI，海外覆盖 Gemini/ChatGPT/Claude/Grok/Perplexity，问题库与指标按市场分开算。当用户说「做 GEO」「生成式引擎优化」「让 AI 推荐我的产品」「AI 搜索里搜不到我们」「GEO 方案/诊断/监测/月报/交付」「给客户做 GEO 服务」，或给一个产品网址要做 AI 可见性优化时使用。GEO 指生成式引擎优化，不是地理信息；不用于传统 SEO 关键词排名、竞价投放或建站。

GEO 全流程服务

从诊断到交付的完整链路，不只是给建议：

抓取 → 体检 → AI答案采样 → 生成工单 → 产出资产 → 报告 → 自动验收 → 客户交付包

核心判断只有一句：GEO 的终点不是排名，是 AI 答案里那句话是不是按你的口径说的。 所以最小单位不是页面，是可被抽取的事实块。

项目目录：/Users/brucejan/geo，每个客户/产品一个 work/<slug>/。

最快的用法：给一个网址

bash

python3 scripts/geo.py new --url https://example.com --market both

九步全自动：建项目 → 抓官网 → 体检 → 自动推导品牌事实/竞品/问题库 → 重跑体检 → AI 答案采样 → 工单与建设蓝图 → 资产与报告 → 三份交付物。

产出 work/<slug>/deliverables/：

文件	回答什么
`1-GEO诊断报告.html`	现在什么样
`2-GEO优化方案.html`	应该改成什么样、为什么
`3-GEO执行方案.html`	谁在什么时候做什么、做到什么算完成

自动推导出来的品牌事实必须人工复核——bootstrap 只从官网正文抽取，抽不到的一律标「待确认」，绝不用常识填充。标了「待确认」的字段（成立时间、工商主体、可具名客户等）需要你补齐或明确不对外说。

两种用法

界面（推荐）

bash

python3 scripts/geo.py ui        # 默认 http://127.0.0.1:8765

全流程都在界面上，不用记命令：新建项目、编辑配置与问题库、写事实卡、一键跑任意步骤或整个周期（带实时日志）、改工单状态、看资产、导入人工采样表、打开交付包。

界面为暗色四段主线（AtlasGEO）：现状（总览 / 引擎表现 / 竞品对比 / 问题库）→ 诊断（差距诊断 / 阵地地图 / 品牌事实库）→ 提升（行动计划 / 内容工作台）→ 成效（效果验收 / 报告与交付），外加设置与三步接入引导。统一口径：GEO 健康分（五项加权，未测项权重归一）、提及率、引用份额、阵地、任务、内容。总览标题是由数据自动生成的一句结论；所有数字来自同一份采样，算不出的显示「未测」，不编数。任务在后台子进程跑，关掉页面也会继续；同一项目同时只允许一个任务，避免抢同一份 audit.json。

只用标准库起服务，前端零外部依赖，断网可用。

命令行

bash

python3 scripts/geo.py serve --slug <项目>

抓取 → 体检 → 采样 → 工单 → 资产 → 报告 → 验收上期 → 打包交付，全做完。产出在 work/<slug>/delivery/<日期>/，可直接发客户。

首次接一个新客户走下面的步骤 0–2 做好底座，之后每期只跑这一条。

先读什么

每次必读：references/method.md（评分口径和所有判据的出处，脚本就是它的代码实现）
排外部信源优先级时必读：references/cn-source-ranking.md （CN-GEO 187,818 条引用实算榜。官网只占 1.37%，这个数字会改变资源分配）
打国内平台：references/cn-platforms.md ｜打海外：references/global-platforms.md
写内容/改页面：references/content-patterns.md
更深方法或原始数据：references/sources.md（含 GEORank / GEOFlow 选型说明）
采样/平台异常先对表：references/known-issues.md（豆包超时、302 国内/海外端点模型目录、 SPA 空壳、llms.txt 范围过滤）

不要凭经验给 GEO 建议。每条建议都要能追到 method.md 里的某条实测数字，说不出依据的别写进方案。

步骤 0 · 建项目

bash

python3 scripts/geo.py init --url <产品官网> --name <品牌名> --market both

--market 取 cn / global / both。

双市场不是"顺带也做海外"，是两套并行的战场——问题库、内容、竞品清单、指标全部分开。纪律见 global-platforms.md 第 5 节。

步骤 1 · 吃透产品，建事实底座

读官网 + 用户给的材料（PDF/PPT/文档/公众号文章都行），补全 geo.json：

brand：规范名、别名和常见错写（品牌消歧的地基）、产品线、行业、目标用户、业务目标
competitors：3–6 个真实竞品，带别名。首期可以留空——先跑一轮"推荐类"问题的采样，从 AI 答案里反推真实竞争集，比主观拍脑袋准得多。双市场时国内外竞品通常不是同一批，都要列全（排名指标是相对这个清单算的，漏掉真实对手会高估名次）

同时写 work/<slug>/content/facts.md 品牌事实卡，模板见 content-patterns.md 第 1 节。 每条事实标证据等级 A–E，没来源的标"待确认"，不许编。 这份文件是后面所有资产生成的输入——llms.txt、JSON-LD、定义块都从它来。

材料不够就直接问用户要，别猜产品能力和价格。

步骤 2 · 建问题库

问题库决定采样什么、写什么内容、怎么算指标。按七组各出 3–6 题：推荐 / 比较 / 替代 / 价格 / 风险 / 品牌验证 / 场景。

每题必须标 market，脚本据此路由——中文题不会打到 Perplexity，英文题不会打到豆包：

json

{"id": "q001", "group": "推荐", "market": "cn",     "text": "国内做私域运营的 SaaS 有哪些好用的？"}
{"id": "q101", "group": "推荐", "market": "global", "text": "What's the best CRM for small B2B teams in 2026?"}
{"id": "q900", "group": "品牌验证", "market": "both", "text": "<品牌> 是家什么公司？"}

cn：中文口语问法，不要翻译腔｜ global：英文原生问法，不是机翻
both 只留给品牌验证类。注意这类问题点名了品牌，答案必然复述品牌名，脚本会把它们单独归入「品牌认知」，不混进可见性指标

每期循环（`serve` 自动做完，这里说明每步在干什么）

抓取 + 体检

按六维打分：可抓取性 / 长度 / 结构 / 可抽取块 / 权威信号 / 对题性。读结果时先看四件事：

SPA 空壳页（word_count 接近 0）——国内官网最常见致命伤，AI 抓取器看到的是空白
robots 有没有封 AI 抓取器——封了就什么都别谈
language_coverage 中英是否对等——做海外却没英文原生页直接 P0
block_gap 缺得最多的块——内容工程第一优先级

AI 答案采样

市场	平台	变量	说明
国内	智谱GLM	`ZHIPUAI_API_KEY`	OpenAI 兼容端点，不联网
国内	豆包	`ARK_API_KEY`	火山方舟；联网要在控制台单独开通内容插件，没开通自动降级
国内	DeepSeek	`DEEPSEEK_API_KEY`	不联网，测模型参数化知识里的品牌认知
国内	Kimi	`MOONSHOT_API_KEY`	默认不联网
国内	MiniMax	`MINIMAX_API_KEY`	OpenAI 兼容端点，不联网
海外	Gemini	`GEMINI_API_KEY`	OpenAI 兼容端点，不带 grounding
海外	OpenAI(ChatGPT)	`OPENAI_API_KEY`	Chat Completions 默认不联网
海外	Claude	`ANTHROPIC_API_KEY`	Anthropic 原生协议，不联网
海外	Grok	`XAI_API_KEY`	xAI API，不联网
海外	Perplexity	`PERPLEXITY_API_KEY`	原生联网并返回 citations，海外证据质量最好

Key 放项目根目录 .env（已 gitignore，权限 600），脚本自动加载。

豆包在 302.AI 模式下走「搜索+回答」组合，单题超时读 .env 的 AI302AI_SEARCH_ASK_TIMEOUT （默认 45s）。平台/端点异常先对 references/known-issues.md 对表，别让任务空转。

没有公开联网 API 的——国内纳米AI搜索/百度AI/豆包 App，海外 ChatGPT 网页版/ Claude 网页版——走人工或浏览器采样：

bash

python3 scripts/geo.py sample-sheet --slug <项目>
python3 scripts/geo.py sample-import --slug <项目> --file <采样表>

口径纪律（详见 method.md 第 4 节）：

国内 ≠ 海外，跨市场平均提及率没有解释力，报告里各占一张表
API ≠ 网页端，Web ≠ App（千问 web 与 App 信源只有 24.5% 重合），单独记单独算
没有答案原文/截图/采样环境记录的，不许标成"真实采样"
归因默认写"观察相关"，除非有基线窗口 + 对照 Prompt + 竞品对照
采样频率克制，遵守平台服务条款，不批量滥采、不模拟登录

Show full SKILL.md (216 more words)Show less

生成工单 `plan`

诊断结果 → tasks.json，这是执行状态的单一真相源。每条工单必须有：依据（追到 method.md 哪一条）、负责角色、工作量、时间窗口、验收标准、市场。

七个工作包：实体消歧 / 页面技术 / 内容矩阵 / 标题体系 / 知识库 / 外部证据 / 监测闭环。

优先级顺序（method.md 第 6 节）：门票问题 → 事实错误 → 抽取块缺口 → 高价值问题无承接 → 外部信源（P1，不是 P2） → 长尾扩量。

外部信源是 P1 不是 P2，因为官网只占全库引用 1.37%。官网是事实源不是引用源，把它从 60 分做到 90 分的边际收益，远低于拿下一个榜单站词条。

产出资产 `generate`

到 work/<slug>/assets/，中英分开：

产物	说明
`llms.txt` / `llms.en.txt`	官方事实索引，直接传网站根目录
`jsonld/*.json`	Organization / SoftwareApplication / FAQPage / Article / BreadcrumbList，贴进 `<head>`
`snippets/definition.*.html`	定义块，放首屏口号下方（口号保留，不影响转化）
`snippets/faq.*.html`	FAQ 块，答案必须在静态 HTML 里可见，不要纯 JS 折叠
`outlines/*.md`	每个目标问题一份内容大纲（含标题候选、章节骨架、字数与抽取块要求）
`drafts/*.md`	加 `--draft` 时调用已配 LLM 出全文初稿

分工：结构性资产由代码确定性生成（不会漏 schema 字段）；文章正文由你按 outline 写—— 代码写不出好文案。--draft 的初稿必须人工核实事实后才能发布。

自动验收 `verify`

重抓站点 → 跑 checker → 回写工单状态。这是"服务"和"建议"的分界线：能自动验收的，就不靠人口头说做完了。做完会翻成 done，回归了会翻回 todo。

无法程序判定的（如百科词条是否过审）标「待人工」，确认后手动标记：

bash

python3 scripts/geo.py task --slug <项目> --id T-003 --status done --note "词条已上线"
python3 scripts/geo.py status --slug <项目>     # 进度看板

客户交付 `deliver`

打包到 work/<slug>/delivery/<日期>/：总览 index、诊断报告、执行方案、工单表（HTML + CSV 可导进项目管理工具）、验收表、建设地图、引擎表现 / 竞品对比 （复用 UI 同一套 analytics 口径，逐引擎提及率/引用份额/样本回放 + 竞品×引擎矩阵/失守问题）、 assets 目录、交付说明。

跑完之后你要做的事（脚本做不了的部分）：

读 reports/latest.md 看 delta：均分涨跌、提及率变化、新出现的 P0
看「AI 实际引用的信源域名 Top」和「与全国大盘对照」——这是当期真实的分发地图，用它修正外部信源计划
检查采样答案里有没有事实错误（品牌说错、张冠李戴、价格错、归错行业）→ 立刻 P0
按 outlines/ 写内容，按 plan.md 更新进度

建议节奏：页面体检每周，答案采样每两周或每月（采样有成本，指标本身有噪声，跑太密看不出信号）。

全部命令

命令	作用
`new`	★ 只给一个网址，全自动出三份交付物
`autopilot`	对已建好的项目跑完整引导流程
`bootstrap`	从官网正文推导品牌事实、竞品、问题库
`deliverables`	出三份正式交付物（诊断/优化/执行）
`init`	只建项目骨架，不跑流程
`crawl` / `audit`	抓站 / 六维体检
`sample` / `sample-sheet` / `sample-import`	API 采样 / 导出人工采样表 / 回灌
`plan`	诊断结果 → 带验收标准的工单
`generate`	产出可部署资产（`--draft` 加 LLM 初稿）
`report`	Markdown + 自包含 HTML 报告，含 delta 与大盘对照
`verify`	重抓并自动验收工单（`--no-recrawl` 用现有结果）
`task` / `status`	单条工单状态 / 项目看板
`deliver`	打包客户交付物
`ui`	全流程界面：新建项目、配置、问题库、事实卡、一键运行、工单、资产、交付
`serve`	全流程一条命令
`cycle`	轻量循环（抓取→体检→采样→报告，不含工单与交付）
`list`	所有项目

排期自动执行

用 schedule skill 建定时任务：

每周一早上跑 python3 /Users/brucejan/geo/scripts/geo.py serve --slug <项目>，然后读 work/<项目>/reports/latest.md，如果出现新的 P0、提及率下降超过 10 个百分点，或有工单从 done 回归成 todo，就告诉我。

边界

做：GEO 诊断、方案、工单、内容工程、资产生成、AI 答案监测、验收闭环、客户交付包。

不做：

传统 SEO 关键词排名、外链买卖、竞价投放
建站/改站的实际代码实施（给方案、给可直接用的片段，落地由客户开发做）
绕过平台限制的批量采集、模拟登录、刷量
编造品牌事实、客户案例、价格或资质——查不到就标"待确认"，问用户

不承诺任何平台一定会引用某个页面。GEO 提高的是概率，不是保证。给客户写方案时这句话必须原样写进去。

/Users/brucejan/geo/
├── SKILL.md
├── references/
│   ├── method.md              评分口径与全部判据出处
│   ├── cn-source-ranking.md   国内信源实测榜（CN-GEO 实算，含复算脚本）
│   ├── cn-platforms.md        国内平台适配
│   ├── global-platforms.md    海外平台适配
│   ├── content-patterns.md    可抽取内容模板
│   └── sources.md             资料索引（含 GEORank / GEOFlow 选型）
├── scripts/
│   ├── geo.py       CLI 总入口
│   ├── crawl.py     抓站      audit.py     六维体检
│   ├── sample.py    多平台采样  benchmark.py 与全国大盘对照
│   ├── tasks.py     工单系统    generate.py  资产生成
│   ├── verify.py    自动验收    deliver.py   客户交付包
│   ├── bootstrap.py 自动推导底座  deliverables.py 三份交付物
│   ├── blueprint.py 建设地图
│   ├── dashboard.py 界面后端    ui.html      前端工作台
│   ├── jobs.py      后台任务（子进程 + 实时日志）
│   └── report.py    报告渲染    geolib.py    共用工具
└── work/<slug>/
    ├── geo.json        品牌、竞品、问题库、平台、目标
    ├── content/facts.md 品牌事实卡（所有资产的输入）
    ├── evidence/       抓取快照      audit.json  体检结果
    ├── samples/        采样原始答案   metrics/    每期指标
    ├── tasks.json      工单（执行状态单一真相源）
    ├── assets/         可部署资产
    ├── reports/        每期报告 + latest.md
    ├── verify/         每期验收结果
    ├── delivery/<日期>/ 客户交付包
    ├── history/        历史基线（算 delta）
    └── plan.md         30/60/90 方案

Files

SKILL.md and 62 other files (scripts, references) in skills/geolook of liangdabiao/GEO-Content-Optimizer-Skill.

SKILL.md
.env.example
.gitignore
LICENSE
README.md
docs/302ai-integration-research.md
docs/ARCHITECTURE.md
docs/ScreenShot_2026-08-04_120350_841.png
docs/demo.gif
docs/screenshots/assets.png
docs/screenshots/channels.png
docs/screenshots/competitors.png
docs/screenshots/engines.png
docs/screenshots/facts.png
docs/screenshots/gaps.png
docs/screenshots/overview.png
docs/screenshots/plan.png
docs/screenshots/questions.png
docs/screenshots/settings.png
… and 44 more

Open the folder on GitHubat commit f56124f

Used in 1 other repository

We found 1 copy of this SKILL.md (exact, near-identical or edited) in other folders, from 1 other GitHub owner. This page covers the copy in liangdabiao/GEO-Content-Optimizer-Skill, which our catalogue first saw on October 7, 2026.

aigclink/geolookrepository root

Compare with similar skills

Geo next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Geo compared with similar skills
Skill	Stars	Used in	Tokens	Auto-check	Licence	Repo updated
Geo this skillliangdabiao/GEO-Content-Optimizer-Skill	205	1 repos	~2.3k	Automated safety check: Notes	MIT	2 mo ago
GEO-Claw AI Visibility AgentLeoYeAI/openclaw-master-skills	2.2k	—	~4.7k	Automated safety check: Pass	MIT	2 mo ago
Agent Readiness Auditindranilbanerjee/digital-marketing-pro	859	1 repos	~3.9k	Automated safety check: Pass	MIT	5 days ago
Geo Optimizerhuifer/claude-code-seo	110	—	~650	Automated safety check: Notes	MIT	9 mo ago
Web Contentalinaqi/maggy	707	—	~3.8k	Automated safety check: Pass	MIT	15 days ago
SEO Geo OptimizerNeverSight/learn-skills.dev	216	1 repos	~2.6k	Automated safety check: Pass	MIT	today

Similar skills

GEO-Claw AI Visibility Agent
LeoYeAI/openclaw-master-skills
Runs a brand's AI-search visibility work end to end: diagnosing how AI platforms represent it, repositioning it, producing AI-optimized content and monitoring ongoing mentions.
2.2k GitHub stars~4.7k tokensUpdated 2 mo ago
Marketing & SEOAuto-check passed
Agent Readiness Audit
indranilbanerjee/digital-marketing-pro
Audit whether AI agents and AI crawlers can actually use a site — robots.txt rules per AI crawler token (OpenAI, Anthropic and Perplexity bots, Google-Extended, Applebot-Extended)…
859 GitHub starsUsed in 1 repo~3.9k tokens
Data & AnalyticsAuto-check passed
Geo Optimizer
huifer/claude-code-seo
生成式引擎优化专家，分析和优化内容在 AI 搜索引擎（ChatGPT、Claude、Perplexity、Google SGE）中的可见性和引用率。
110 GitHub stars~650 tokensUpdated 9 mo ago
Marketing & SEOAuto-check: notes
Web Content
alinaqi/maggy
SEO and AI discovery (GEO) - schema, ChatGPT/Perplexity optimization
707 GitHub stars~3.8k tokensUpdated 15 days ago
Marketing & SEOAuto-check passed
SEO Geo Optimizer
NeverSight/learn-skills.dev
Comprehensive SEO/GEO/AEO analysis toolkit for optimizing content visibility across traditional search engines (Google, Bing), AI platforms (ChatGPT, Perplexity, Claude, Gemini, Grokipedia), answer…
216 GitHub starsUsed in 1 repo~2.6k tokens
Marketing & SEOAuto-check passed
Money SEO
iamzifei/show-me-the-money
SEO and GEO (Generative Engine Optimization) for organic traffic and AI search visibility.
1k GitHub stars~4.4k tokensUpdated 1 mo ago
Marketing & SEOAuto-check passed

Questions about Geo

What does Geo do?

When should I use Geo?

Geo fits situations like: tasks that involve AI search optimization; tasks that involve Web search; tasks that involve Schema markup.

How do I install Geo in Claude Code?

Run `npx skills add liangdabiao/GEO-Content-Optimizer-Skill --skill geo -a claude-code`. Or copy the skill folder (skills/geolook in liangdabiao/GEO-Content-Optimizer-Skill) into .claude/skills/geo in your project. Claude Code loads it when a task matches its description.

How do I install Geo in Codex?

Run `npx skills add liangdabiao/GEO-Content-Optimizer-Skill --skill geo -a codex`. Or copy the skill folder (skills/geolook in liangdabiao/GEO-Content-Optimizer-Skill) into .agents/skills/geo in your project. Codex loads it when a task matches its description.

Can I use Geo in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add liangdabiao/GEO-Content-Optimizer-Skill --skill geo -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/geo, .gemini/skills/geo, .github/skills/geo and .opencode/skills/geo in your project.

What does Geo need to run?

Going by SKILL.md and its folder, Geo needs the command-line tools its instructions call (python3) and credentials named ZHIPUAI_API_KEY, ARK_API_KEY, DEEPSEEK_API_KEY and MOONSHOT_API_KEY. Our summary lists: Python 3; A credential in ZHIPUAI_API_KEY; A credential in ARK_API_KEY.

Does Geo access the network?

SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.

Is Geo safe to install?

Our automated static check of SKILL.md found notes only (mentions a .env file), nothing it rates as a warning. It is not a guarantee. The check reads SKILL.md only: the scripts in the folder are not scanned, so read them before running anything.

What licence does Geo use?

Geo is published under the MIT licence (from the LICENSE file in the skill folder). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Geo use?

About 2.3k tokens (SKILL.md is roughly 9.2k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full. Its references folder adds about 10k tokens, read only when the agent opens those files.

What are the alternatives to Geo?

Skills that share tags, products or a category with Geo: GEO-Claw AI Visibility Agent (LeoYeAI/openclaw-master-skills, 2.2k stars), Agent Readiness Audit (indranilbanerjee/digital-marketing-pro, 859 stars), Geo Optimizer (huifer/claude-code-seo, 110 stars) and Web Content (alinaqi/maggy, 707 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Geo?

liangdabiao (a GitHub user) maintains it in liangdabiao/GEO-Content-Optimizer-Skill, which has 205 GitHub stars. The repository holds 3 skills in this directory. The repository was last updated on August 5, 2026.

Source: liangdabiao/GEO-Content-Optimizer-Skill on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.

Install Geo

Facts

At a glance

What it does

When your agent uses it

Example prompts

Requirements

Workflow steps

What it can do on your machine

Tool permissions

Runs code

Network

Credentials

Context cost

Safety

SKILL.md

GEO 全流程服务

最快的用法：给一个网址

两种用法

界面（推荐）

命令行

先读什么

步骤 0 · 建项目

步骤 1 · 吃透产品，建事实底座

步骤 2 · 建问题库

每期循环（serve 自动做完，这里说明每步在干什么）

抓取 + 体检

AI 答案采样

生成工单 plan

产出资产 generate

自动验收 verify

客户交付 deliver

全部命令

排期自动执行

边界

目录

Files

Used in 1 other repository

Compare with similar skills

Similar skills

More from liangdabiao/GEO-Content-Optimizer-Skill

Related

Questions about Geo

每期循环（`serve` 自动做完，这里说明每步在干什么）

生成工单 `plan`

产出资产 `generate`

自动验收 `verify`

客户交付 `deliver`