Token导航 LogoToken导航

Agent Skill · 研究检索

go-mode(走模式)

go-mode 用于查找、检索和筛选相关信息,适合在 Codex、Claude、Cursor、Gemini CLI 中需要根据关键词、任务场景或来源线索快速定位候选结果时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

go-mode 是研究检索类 Agent Skill,可用宿主为 Codex、Claude、Cursor、Gemini CLI,通过GitHub 仓库获取,权限为执行命令;许可证信息还不明确,第三方扫描有提醒项,按安装量在研究检索类的 73,675 个 Skill 里排第 37,901。

总安装 1,648周安装 68GitHub Stars 261下载 539核验 2026-05-01

安装说明

本站只整理中文说明,不托管安装包,也不代用户安装。
GitHub
许可证
unknown

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:go-mode(走模式)
来源仓库:https://github.com/brianrwagner/ai-marketing-claude-code-skills
仓库路径:skills/go-mode
安装命令:
npx skills add https://github.com/brianrwagner/ai-marketing-claude-code-skills --skill go-mode
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 npx skills 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

skills.shnpx skills
npx skills add https://github.com/brianrwagner/ai-marketing-claude-code-skills --skill go-mode

这个 Skill 是做什么的

研究检索类 Skill 主要解决「找资料、核对事实、把来源整理清楚」的问题,通常会让助手先检索再总结。

用于查找、检索和筛选相关信息,支持关键词匹配和线索定位。适合在需要快速定位候选结果的任务场景中使用。可结合任务目标提供相关技术资料或代码片段参考。输出结果需人工复核,避免直接依赖工具返回结论。安装前建议确认来源仓库维护状态和权限范围。go-mode 属于研究检索类 Skill,可作为该场景下的辅助能力补充。

来源仓库内路径:skills/go-mode

上手步骤

4 步
  1. 1

    确认自己的宿主工具

    公开资料标注支持 Codex、Claude、Cursor、Gemini CLI;先确认自己用的是其中之一,再按对应工具的 Skills 目录放置。

  2. 2

    拿到安装命令

    可以直接复制的命令是 npx skills add https://github.com/brianrwagner/ai-marketing-claude-code-skills -…;命令里的占位符换成自己的路径或密钥再执行。

  3. 3

    安装后重启宿主

    命令执行完要完全退出宿主工具再打开,让新装的 Skill 被重新扫描到,否则列表里可能看不到。

  4. 4

    按最小权限试用

    它需要执行命令,先在测试目录或测试账号里跑,确认没问题再放到正式项目。

常见问题

6 条

go-mode 怎么安装?

在终端执行 npx skills add https://github.com/brianrwagner/ai-marketing-claude-code-skills --skill go-…。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

go-mode 支持哪些 AI 工具?

公开资料标注支持 Codex、Claude、Cursor、Gemini CLI。其他工具只要能读取本地 Skill 目录,一般也能手动放进去,但要自己确认触发方式是否一致。

go-mode 需要什么权限?

权限级别是执行命令。安装流程涉及命令执行,可能通过 npx skills add https://github.com/brianrwagner/ai-marketing-claude-code-skills --skill go-mode 联网下载 Skill 或依赖。用户安装前应确认命令来源、仓库内容和执行环境。

go-mode 是免费的吗?

公开资料没有标注明确许可证,收费和商用限制都无法确认。正式使用前请到来源仓库核对授权说明。

go-mode 用的人多吗?

公开数据是累计安装 1,648 次,近一周安装 68 次,下载 539 次,GitHub 261 星。这些数字来自第三方目录统计,只能当作热度参考,不代表质量或安全性。

go-mode 安全吗?

第三方扫描有提醒项,说明存在需要人工确认的行为。安装前请打开源码看清它读了什么、联网到哪里。

值得选的地方和要先想清楚的地方

值得选的地方

  • 同时适配 4 种工具,换客户端不用重新找替代品。

先想清楚的地方

  • 需要执行命令,装到常用环境前要先评估影响面。
  • 许可证没有明确标注,能不能商用、能不能再分发都不确定。
  • 第三方扫描存在提醒项或未通过项,需要人工逐条确认。
  • 第三方目录整理,版本更新和行为细节以来源仓库为准。

适不适合你

更适合

  • 已经在用 Codex,想直接扩展它能力的人。
  • 需要处理研究检索类任务、又不想每次都手写提示词的人。

不太适合

  • 不能接受 Skill 在你机器上执行命令的人。
  • 需要明确授权才能用到公司项目里的场景。
  • 需要厂商级技术支持和 SLA 的生产系统。

研究检索类 Skill 对比

6 个同类
Skill总安装GitHub 星权限许可证扫描
find-skills

寻找技能 · Codex、Claude

31,199,81316,486操作浏览器MIT有提醒
self-improving-agent

自我改进 Agent · OpenClaw

16,464,0023,396接触敏感数据MIT-0通过
soultrace

灵魂追踪 · Codex、Claude

7,588,94412需要联网未明确有提醒
self-improving

自我完善 · OpenClaw

6,829,2841,046需要联网MIT-0通过
microsoft-foundry

微软代工厂 · Codex、Claude

6,614,144764调用外部服务未明确有提醒
go-mode

走模式 · 本页

1,648261执行命令未明确有提醒

按安装量取同分类前几名;数字来自第三方目录统计,只作热度参考,不代表质量结论。

SKILL.md 原文

🎯 Go Mode — Autonomous Goal Execution

Give me a goal. I'll plan it, confirm with you, execute it, and report back. You steer — I drive.

Mode

Detect from context or ask: *"Just do it, plan first, or plan + phase approvals?"*

ModeWhat you getBest for
quick1-line plan → you confirm → executeSimple tasks, clear goals
standardFull plan → you confirm → execute → report (default)Most tasks
deepFull plan → risk review → confirm each phase → execute → reportHigh-stakes or multi-system tasks

Default: standard — use quick for simple, clear goals. Use deep when mistakes would be expensive to undo.


How It Works

GOAL → PLAN → CONFIRM → EXECUTE → REPORT

Phase 1: PLAN

When given a goal, break it down:

  1. Parse the goal — What's the desired outcome? What does "done" look like?
  2. Break into steps — Ordered task list, each step concrete and actionable
  3. Identify tools — Which skills, APIs, agents, or CLI tools are needed?
  4. Estimate effort — Time per step, total duration, API costs if applicable
  5. Flag risks — What could go wrong? What needs human approval?

Output a structured plan:

## 🎯 Goal: [restated goal]

### Definition of Done
[What success looks like]

### Plan
| # | Step | Tool/Skill | Est. Time | Cost | Risk |
|---|------|-----------|-----------|------|------|
| 1 | ... | ... | ... | ... | ... |

### Total Estimate
- **Time:** X minutes
- **API Cost:** ~$X.XX
- **Human Checkpoints:** [list]

### Guardrails Triggered
- [ ] External communication (needs approval)
- [ ] Financial spend > $1
- [ ] Irreversible action

Phase 2: CONFIRM

Present the plan and wait for approval:

  • "Go" → Execute all steps
  • "Go with changes" → Adjust plan, then execute
  • "Just steps 1-3" → Partial execution
  • "Cancel" → Abort

Never skip confirmation. This is the human's steering wheel.

Phase 3: EXECUTE

Run each step sequentially:

  1. Announce the current step: "Step 2/5: Researching competitor pricing..."
  2. Execute using the identified tool/skill
  3. Checkpoint after each major step — brief status update
  4. Pause if:

- A guardrail is triggered (external action, spend, irreversible) - Something unexpected happens - A decision point requires human judgment

  1. Adapt — If a step fails, try alternatives before escalating

Phase 4: REPORT

When all steps complete:

## ✅ Goal Complete: [goal]

### What Was Done
- Step 1: [result]
- Step 2: [result]
- ...

### Outputs
- [List of files, links, artifacts created]

### What Was Learned
- [Insights discovered during execution]

### Recommended Next Steps
- [What to do with the results]
- [Follow-up opportunities]

### Stats
- Total time: Xm
- API calls: X
- Est. cost: $X.XX

Guardrails

Always Ask Before:

  • ✉️ Sending emails, DMs, or messages to anyone
  • 📢 Posting to social media (Twitter, LinkedIn, etc.)
  • 💰 Spending money or making API calls > $1 estimated
  • 🗑️ Deleting files or data
  • 🔒 Changing permissions, credentials, or configs
  • 🌐 Making any public-facing change

Auto-Proceed On:

  • ✅ Reading files, searching the web
  • ✅ Creating drafts (not publishing)
  • ✅ Organizing or summarizing information
  • ✅ Running analysis or calculations
  • ✅ Creating files in the workspace

Budget Caps

  • Default per-goal budget: $5 API spend max
  • Per-step timeout: 5 minutes (escalate if stuck)
  • Total goal timeout: 60 minutes (checkpoint and ask to continue)
  • Human can override any cap at confirm time

Available Tools & Skills Reference

When planning, draw from this toolkit:

Research & Information

ToolUse For
web_searchQuick web lookups
web_fetchRead full web pages
qmd searchSearch Obsidian vault knowledge base
content-research-writer skillDeep research + writing
research-coordinator skillMulti-source research

Content Creation

ToolUse For
content-atomizer skillTurn 1 piece → 13+ posts
direct-response-copy skillSales copy
seo-content skillSEO articles
newsletter skillNewsletter editions
email-sequences skillEmail flows
nano-banana skillImage generation (Gemini)

Marketing & Strategy

ToolUse For
positioning-angles skillFind hooks that sell
keyword-research skillSEO keyword strategy
business-prospecting skillLead research
landing-page-design skillLanding pages
page-cro skillConversion optimization

Communication

ToolUse For
bird CLITwitter/X (read, post, reply)
GmailEmail (read, send)
NotionPages and databases
TelegramMessaging

Development

ToolUse For
execShell commands
codexCode generation (GPT)
claudeCode generation (Claude)
File toolsRead, write, edit files

Example Goals

1. Competitor Analysis → Comparison Page

Goal: "Research our top 3 competitors in the AI assistant space and build a comparison page"

Plan:
1. Identify top 3 competitors (web search) — 5min
2. Research each: pricing, features, reviews — 15min
3. Build comparison matrix — 10min
4. Write comparison page copy — 15min
5. Create visual comparison table — 5min
Total: ~50min, ~$0.50 API cost

2. Content Repurposing Pipeline

Goal: "Take my latest blog post and turn it into a week of social content"

Plan:
1. Read and analyze the blog post — 2min
2. Extract key themes and quotes — 5min
3. Generate 5 Twitter threads — 15min
4. Generate 5 LinkedIn posts — 15min
5. Create 3 image prompts + generate visuals — 10min
6. Build content calendar — 5min
Total: ~52min, ~$1.00 API cost

3. Lead Research Sprint

Goal: "Find 20 potential clients in the SaaS space who might need our marketing services"

Plan:
1. Define ideal client profile — 5min
2. Search for SaaS companies (web) — 15min
3. Research each company's marketing gaps — 20min
4. Score and rank prospects — 10min
5. Build outreach-ready prospect list — 10min
6. Draft personalized intro messages [NEEDS APPROVAL] — 15min
Total: ~75min, ~$0.75 API cost

4. SEO Content Sprint

Goal: "Create 3 SEO-optimized blog posts for our target keywords"

Plan:
1. Review target keyword list — 2min
2. Research top-ranking content for each keyword — 15min
3. Create outlines using SEO skill — 10min
4. Write article 1 — 15min
5. Write article 2 — 15min
6. Write article 3 — 15min
7. Add internal links and meta descriptions — 10min
Total: ~82min, ~$2.00 API cost

5. Launch Prep Checklist

Goal: "Prepare everything needed to launch our new product next Tuesday"

Plan:
1. Audit what exists (landing page, emails, social) — 10min
2. Identify gaps — 5min
3. Write launch email sequence (3 emails) — 20min
4. Create social media posts (Twitter, LinkedIn) — 15min
5. Generate launch graphics — 10min
6. Build launch day timeline — 5min
7. Draft press/outreach messages [NEEDS APPROVAL] — 15min
Total: ~80min, ~$2.50 API cost

6. Weekly Review & Planning

Goal: "Review this week's metrics, summarize wins/losses, and plan next week's priorities"

Plan:
1. Pull metrics from available sources — 10min
2. Summarize key wins — 5min
3. Identify what didn't work — 5min
4. Review upcoming calendar — 5min
5. Propose next week's top 3 priorities — 10min
6. Create actionable task list — 5min
Total: ~40min, ~$0.25 API cost

Usage

Just tell the agent your goal in natural language:

"Take the wheel: Research the top 5 AI newsletter tools, compare them, and recommend the best one for a solopreneur"
"Take the wheel: Build a complete email welcome sequence for new subscribers — 5 emails over 2 weeks"
"Take the wheel: Audit our Twitter presence and create a 30-day content strategy"

The agent will plan, confirm, execute, and report. You stay in control at every checkpoint.

Principles

  1. Transparency — Always show the plan before executing
  2. Safety — Never take external actions without approval
  3. Efficiency — Use the cheapest/fastest tool for each step
  4. Resilience — Try alternatives before giving up
  5. Accountability — Report everything that was done
  6. Respect time — Estimate honestly, checkpoint if running long

平台分布

Codex

35.03%
按下载量换算189

Claude

33.17%
按下载量换算179

Cursor

19.74%
按下载量换算106

Gemini CLI

9.73%
按下载量换算52

安全审计

Gen Agent Trust Hub

可疑

Socket

通过

Snyk

可疑

权限和风险

执行命令

安装流程涉及命令执行,可能通过 npx skills add https://github.com/brianrwagner/ai-marketing-claude-code-skills --skill go-mode 联网下载 Skill 或依赖。用户安装前应确认命令来源、仓库内容和执行环境。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。来源安全扫描存在提醒或未通过项,安装前请自行复核。目前只有一个公开来源,建议同时核对源仓库。

继续浏览同类 Skills