Token导航 LogoToken导航

Agent Skill · 研究检索

setup-memory(设置记忆)

setup-memory 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。

setup-memory 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,权限为只读;许可证为 MIT,第三方扫描已通过,按安装量在研究检索类的 73,674 个 Skill 里排第 59,464。

总安装 388周安装 16GitHub Stars 2下载 127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
MIT

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:setup-memory(设置记忆)
来源仓库:https://github.com/abdullahbeam/nexus-design-abdullah
仓库路径:skills/setup-memory
安装命令:
npx skills add https://github.com/abdullahbeam/nexus-design-abdullah --skill setup-memory
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。不同来源提供的安装方式可能略有差异;本站展示可直接复制的安装命令,安装前请核对来源页面。

skills.shnpx skills
npx skills add https://github.com/abdullahbeam/nexus-design-abdullah --skill setup-memory

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。

提供记忆存储与上下文管理的基础设施支持。适用于 Codex、Claude、Cursor、Gemini CLI 中需要持久化会话状态的应用。通过 npx skills add 命令从 nexus-design-abdullah 仓库安装。涉及数据读写操作,应确认存储路径权限及数据加密要求。

注意内存使用效率,尤其在长时间运行任务中需监控资源消耗。

来源仓库内路径:skills/setup-memory

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/abdullahbeam/nexus-design-abdullah --skill set…;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

    它标注为只读,先用只读任务试一次,确认行为符合预期再扩大使用范围。

常见问题

6 条

setup-memory 怎么安装?

在终端执行 npx skills add https://github.com/abdullahbeam/nexus-design-abdullah --skill setup-memory。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

setup-memory 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。

setup-memory 需要什么权限?

权限级别是只读。该 Skill 主要提供规则、说明或参考内容,本身偏只读;真正读写文件、联网或执行命令仍取决于宿主 Agent 的任务。

setup-memory 是免费的吗?

来源标注的许可证是 MIT。本站不代收任何费用,也不托管安装包;能否商用、要不要署名,以许可证原文为准。

setup-memory 用的人多吗?

公开数据是累计安装 388 次,近一周安装 16 次,下载 127 次,GitHub 2 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。

setup-memory 安全吗?

第三方扫描结果为通过,但这只覆盖已知的静态检查项,不能替代自己读一遍源码和安装命令。

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。
  • 3 个目录同时收录,说明这个 Skill 流传比较广。
  • 许可证明确为 MIT,商用和分发边界好判断。
  • 第三方扫描已通过,常见的高风险行为没有命中。

先想清楚的地方

  • 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。
  • 需要许可证清楚、并且有第三方扫描结果的团队项目。

不太适合

  • 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
setup-memory

设置记忆 · 本页

3882只读MIT通过

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

🎯 AI Proactive Triggering (ONBOARDING SKILL - CRITICAL PRIORITY)

This is the FIRST onboarding skill with CRITICAL priority. Suggest early and often.

When to Proactively Suggest (AI MUST check user-config.yaml)

Check learning_tracker.completed.setup_goals AND goals.md for smart_default: true. If not personalized:

PROACTIVELY SUGGEST when user:

  1. FIRST SESSION: Always suggest during first interaction if goals not set
  2. Asks for help with work that would benefit from context (AI can personalize better after setup)
  3. Uses Nexus for meaningful work but hasn't personalized yet
  4. Mentions their role, job, or what they do
  5. Expresses frustration that AI doesn't understand their context
  6. At menu display when goals_personalized: false - PROMINENTLY highlight

Suggestion Pattern (first session):

💡 Welcome to Nexus! I'm currently using smart defaults. To help you most
effectively, I'd love to learn about:
- Your role and work context
- Your goals (short-term and long-term)
- Your preferences

This takes about 8 minutes and dramatically improves our collaboration.
Say 'setup goals' to personalize, or continue with defaults.

Suggestion Pattern (returning user, still on defaults):

💡 I notice you're still using Nexus defaults. Personalizing takes 8 minutes
and helps me understand your work context, goals, and preferences.

Ready to 'setup goals'? (This is a one-time setup that improves every session)

Menu Integration: When displaying menu with goals_personalized: false:

🧠 MEMORY
   ⚠️ Not personalized ▸ 'setup goals' (8 min, highly recommended)

DO NOT suggest if:

  • learning_tracker.completed.setup_goals: true
  • goals.md no longer has smart_default: true
  • User explicitly declined personalization multiple times

Setup Goals

Guide user through goal definition and system personalization.

Purpose

Transform smart default templates into meaningful, personalized context that improves AI collaboration quality. Captures user's role, short-term goals (3 months), long-term vision (1-3 years), and work preferences.

Time Estimate: 8-10 minutes


Workflow

Step 1: Welcome & Language

Display:

━━━ SETUP GOALS ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Let's personalize Nexus to understand your work context.
This takes about 8-10 minutes and improves AI collaboration.
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

Ask: "What language would you like me to use? (English is default)"

Action: Store preference in user-config.yaml, switch all communication.


Step 2: Role Discovery

Ask: "What do you do? Tell me about your current role or situation."

AI Suggestion Pattern: Listen, then offer 2-3 refined versions. Let user pick or refine.

Store: Update ## Current Role in goals.md


Step 3: Short-Term Goal

Ask: "What's the ONE thing you want to achieve in the next 3 months?"

Help make it specific and measurable. Capture:

  • The goal itself
  • Why it matters (motivation)
  • 2-3 success metrics

Store: Update ## Short-Term Goal (3 months) in goals.md


Step 4: Long-Term Vision

Ask: "Where do you want to be in 1-3 years?"

Connect to short-term goal to show trajectory.

Store: Update ## Long-Term Vision (1-3 years) in goals.md


Step 5: Work Preferences

Quick questions:

  1. "When do you do best work?" (morning/afternoon/evening)
  2. "Typical session length?" (30min, 1hr, 2hrs+)
  3. "What types of work?" (writing, coding, research, planning...)

Store: Update ## Work Style & Preferences in goals.md


Step 6: Finalize

Actions (MUST complete all):

  1. Remove smart_default: true from goals.md YAML frontmatter (if present)
  2. Update Last Updated timestamp in goals.md
  3. Mark skill complete in user-config.yaml: learning_tracker: completed: setup_goals: true # ADD THIS LINE
  4. Update language in user-config.yaml (if user specified): user_preferences: language: "{user's language}"
  5. Display completion: ✅ Setup Goals Complete! Captured: • Your role and work context • Short-term goal (3 months) with success metrics • Long-term vision (1-3 years) • Work style preferences Nexus now understands you. Context loads every session. Next steps: • 'setup workspace' - Organize your folders • 'learn projects' - Understand project workflow • 'create project' - Start working

Step 7: Close-Session Practice

Display:

━━━ IMPORTANT HABIT ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
When done working, always say "done" or "close session".
This saves progress and helps me remember context.
Let's practice - say "done" now!
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

Wait for "done", then trigger close-session.


Success Criteria

  • Language preference captured in user-config.yaml
  • Role clearly defined in goals.md
  • Short-term goal specific and measurable
  • Success metrics defined
  • Long-term vision captured
  • smart_default: true removed from goals.md
  • learning_tracker.completed.setup_goals: true in user-config.yaml
  • User practiced close-session

平台分布

Claude Code

31.1%
按下载量换算39

Antigravity

21.96%
按下载量换算28

windsurf

17%
按下载量换算22

Codex

11.62%
按下载量换算15

OpenCode

8.6%
按下载量换算11

Gemini CLI

3.27%
按下载量换算4

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

通过

权限和风险

只读

该 Skill 主要提供规则、说明或参考内容,本身偏只读;真正读写文件、联网或执行命令仍取决于宿主 Agent 的任务。

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。

继续浏览同类 Skills