Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add redfox-data/redfox-community-dsh --skill geo-analyzergit clone --depth 1 https://github.com/redfox-data/redfox-community-dshWrote 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.
[](https://agentmods.dev/skills/redfox-data/redfox-community-dsh/geo-analyzer)<a href="https://agentmods.dev/skills/redfox-data/redfox-community-dsh/geo-analyzer"><img src="https://agentmods.dev/badge/skills/redfox-data/redfox-community-dsh/geo-analyzer/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.
<a href="https://agentmods.dev/skills/redfox-data/redfox-community-dsh/geo-analyzer"><img src="https://agentmods.dev/badge/skills/redfox-data/redfox-community-dsh/geo-analyzer.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00136 | $0.03277 |
| Opus 5 | $0.00068 | $0.01639 |
| Sonnet 5 | $0.00027 | $0.00655 |
| Haiku 4.5 | $0.00014 | $0.00328 |
Grade A, and why
geo-analyzer 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 9d 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.
How it starts
The opening of the file, as written. The whole thing — 295 lines — stays where its author put it; the contents beside it link to each section on GitHub.
品牌GEO分析
简介
帮助品牌了解自己在 AI 搜索引擎（豆包、Kimi、DeepSeek）中的表现。系统自动向三个 AI 平台提出相同问题，分析品牌出现率、情绪倾向、信源引用、竞品对比，生成可视化 HTML 报告。
鉴权
前往 红狐hub 获取 API Key，通过以下方式配置：
# 方式一：配置文件
{ "env": { "REDFOX_API_KEY": "ak_xxxx..." } }
# 方式二：终端环境变量
export REDFOX_API_KEY="ak_xxxx..."
依赖
pip3 install requests
完整工作流
Step 0: 输入收集
向用户收集以下信息（使用 AskUserQuestion 或直接从用户消息中提取）：
必填信息：
- 品牌名称: 用户要分析的品牌名（如"元气森林"、"大疆"、"蔚来"）
- 品类/行业: 品牌所属品类（如"无糖饮料"、"无人机"、"新能源汽车"）
可选信息：
- 品牌别名: 品牌的其他常见称呼（如"小红书"的别名"RED"、"小红书App"）
- 竞品列表: 用户已知竞品（如"可口可乐"、"百事可乐"）
- 自定义问题列表: 如果用户已有问题列表，直接使用，跳过 Step 1
关键规则：
- 如果用户只提供了品牌名没有品类，必须追问品类
- 如果用户提供了问题列表，跳过 Step 1 直接进入 Step 2
- 竞品和别名可以为空
Step 1: 问题生成（用户未提供问题时）
如果用户没有提供问题列表，需要生成 5 个热门问题。
1.1 搜索品类热度
调用任一 websearch skill 搜索品类相关信息：
python3 ~/.agents/skills/doubao-websearch/scripts/doubao_search.py "{品类} 消费者最关心的问题"
python3 ~/.agents/skills/kimi-websearch/scripts/kimi_search.py "{品类} 品牌推荐 常见问题"
1.2 AI 生成 5 个问题
结合搜索结果和品类知识，生成 5 个用户最可能在 AI 搜索引擎中提问的问题。
问题类型必须覆盖以下四类（每类至少 2 个）：
| 类型 | 示例 | 说明 |
|---|---|---|
| 推荐类 | "{品类}哪个品牌好？"、"推荐几款好用的{品类}" | 测试品牌是否进入推荐列表 |
| 对比类 | "{品牌A}和{品牌B}哪个好？"、"{品类}品牌对比" | 测试品牌在直接对比中的表现 |
| 评价类 | "{品牌}怎么样？"、"{品牌}值得买吗？" | 测试品牌单独评价的倾向 |
| 场景类 | "{场景}用什么{品类}好？"、"新手适合用哪个{品类}？" | 测试品牌在特定场景下的可见度 |
问题生成规则：
- 问题必须是用户真实可能搜索的自然语言
- 不要在问题中直接包含用户品牌名（推荐类和场景类），除非是评价类问题
- 评价类问题中可以包含品牌名
- 问题长度控制在 10-30 字
- 不要生成过于相似的问题
1.3 确认问题列表（必须执行）
将生成的 5 个问题展示给用户，使用 AskUserQuestion 询问确认。必须等待用户明确同意后，才能进入 Step 2 批量搜索。如果用户要求修改，重新调整问题列表并再次确认。
Step 2: 批量搜索
将 5 个问题同时提交到 3 个 AI 平台进行联网搜索。
python3 scripts/geo_search.py --queries '["问题1","问题2",...,"问题10"]' --platforms doubao,kimi,deepseek
脚本自动完成：
- 批量提交 15 个搜索任务（3平台 x 5问题）
- 并行轮询所有任务，每 5 秒检查一次
- 最长等待 8 分钟
- 输出
output/search_results.json
输出文件结构：
{
"queries": ["问题1", "问题2", ...],
"platforms": ["doubao", "kimi", "deepseek"],
"total_tasks": 30,
"completed": 28,
"failed": 2,
"results": [
{
"question": "问题1",
"query_index": 0,
"platform": "doubao",
"content": "AI回答全文...",
"sources": [{"title": "...", "url": "...", "domain": "..."}],
"status": "completed"
}
]
}
What ships with it
10 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
- README.en.md 4.5 KB
- README.md 4.1 KB
- references/geo-metrics.md 5.0 KB
- scripts/geo_analyze.py 12 KB runs code
- scripts/geo_report.py 3.2 KB runs code
- scripts/geo_search.py 7.9 KB runs code
- scripts/lib/__init__.py 0 B runs code
- scripts/lib/analyzer.py 21 KB runs code
- scripts/lib/platforms.py 8.6 KB runs code
- scripts/lib/report_template.py 50 KB runs code
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.
- 9d ago First seen · 295 lines · 136 tokens per session scan A ac893f98ba0f
geo-analyzer is a skill published in the GitHub repository redfox-data/redfox-community-dsh (5 stars, last pushed today), licensed MIT. It adds 136 tokens to every session and 3,277 once invoked, about $0.0007 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.
Other skills, from other repositories
dsh-hooks-authoring
Instructions for writing and debugging dsh-hooks-plugin scripts, which run during tool or session events such as before a tool call or after it finishes. These scripts return JSON decisions such as allowing, asking about, or denying an action.
web-artifact-designer
A design workflow for producing self-contained HTML or SVG files that open directly in a browser. It covers visual work such as posters, infographics, landing pages, charts, banners, cards, and interface mockups.
dsh-sdk-upgrade
Safely select and install a compatible official @deepseek-ai SDK release for dsh plugin projects (dsh-web, dsh-trading, and similar monorepos) from npm using an isolated worktree, explicit cohort review, CI-equivalent validation, and controlled rollout — including syncing the project's declared DSH host-version floor…
dsh-web-pet-developer
Create a pet for the dsh-pet plugin and integrate it into the dsh web GUI — author a v2 pet.json manifest plus an 8-column x 9-row atlas per the Codex/hatch-pet contract (live2d pets, voice packs and status decorations included), drop it into the pet-center user directory or contribute it as a built-in asset under…
company-research-method
A documented process for researching a company using supplied files and selected public sources. It produces a company overview with business details, sourced financial figures, and stated risks.
dsh-plugin-guide
Use when developing, reviewing, packaging, debugging, or answering questions about DeepSeek Harness (DSH) plugins — the plugin-based agent harness on vendored Cordis. Applies the official plugin-development constraints (plugin contract, cordis.yml layers, services/events/effects, tool DSL, bundles/profiles) backed by…