Token导航 LogoToken导航

Agent Skill · 运维和基础设施

memory-retrieval-learning(记忆检索学习)

用于搭建或维护带检索增强的 RAG 工作流,适合让 Agent 处理知识库问答、向量检索、来源引用和事实核查。它可以辅助整理数据接入、Embedding、向量库、召回参数和回答生成流程。使用时需要确认数据来源、更新频率、召回阈值和引用展示方式,避免把未命中的资料或过期内容包装成确定事实。

memory-retrieval-learning 是运维和基础设施类 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,按安装量在运维和基础设施类的 2,449 个 Skill 里排第 359。

总安装 5,965周安装 239GitHub Stars 85下载 1,931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:memory-retrieval-learning(记忆检索学习)
来源仓库:https://github.com/lyndonkl/claude
仓库路径:skills/memory-retrieval-learning
安装命令:
npx skills add https://github.com/lyndonkl/claude --skill memory-retrieval-learning
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/lyndonkl/claude --skill memory-retrieval-learning

这个 Skill 是做什么的

运维和基础设施类 Skill 处理部署、监控、日志和云资源,多数需要较高的执行权限。

memory-retrieval-learning 用于搭建或维护带检索增强的 RAG 工作流,适合让 Agent 处理知识库问答、向量检索、来源引用和事实核查。它可以辅助整理数据接入、Embedding、向量库、召回参数和回答生成流程。

使用时需要确认数据来源、更新频率、召回阈值和引用展示方式,避免把未命中的资料或过期内容包装成确定事实。该技能适用于运维和基础设施类别,主要用于 RAG 系统的构建和维护。使用时需特别注意数据新鲜度和引用准确性要求。

来源仓库内路径:skills/memory-retrieval-learning

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/lyndonkl/claude --skill memory-retrieval-learn…;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

memory-retrieval-learning 怎么安装?

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

memory-retrieval-learning 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。

memory-retrieval-learning 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。

memory-retrieval-learning 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。

memory-retrieval-learning 用的人多吗?

公开数据是累计安装 5,965 次,近一周安装 239 次,下载 1,931 次,GitHub 85 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。

memory-retrieval-learning 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。
  • 需要处理运维和基础设施类任务、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。

运维和基础设施类 Skill 对比

6 个同类
Skill总安装GitHub 星权限许可证扫描
brainstorming

头脑风暴 · Codex、Claude

3,148,776173,131操作浏览器MIT通过
copywriting

文案写作 · Codex、Claude

1,996,72425,583需要联网MIT通过
openclaw-secure-linux-cloud

OpenClaw secure linux cloud 命令行 · Codex、Claude

1,853,08840接触敏感数据未明确未通过
xget

xget 命令行 · Codex、Claude

1,799,28040执行命令未明确未通过
requesting-code-review

代码审查请求 · Codex、Claude

1,590,144173,103权限未标注MIT通过
memory-retrieval-learning

记忆检索学习 · 本页

5,96585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

Memory, Retrieval & Learning

Table of Contents

Workflow

Copy this checklist and track your progress:

Learning Plan Progress:
- [ ] Step 1: Define learning goals and timeline
- [ ] Step 2: Break down material and create schedule
- [ ] Step 3: Design retrieval practice methods
- [ ] Step 4: Execute daily learning sessions
- [ ] Step 5: Track progress and adjust

Step 1: Define learning goals and timeline

Clarify what needs to be learned, by when, and how much time is available daily. Identify success criteria (pass exam, demonstrate skill, etc). Use resources/template.md to structure your plan.

Step 2: Break down material and create schedule

Chunk material into learnable units. Calculate spaced repetition schedule based on timeline. Plan initial learning + review cycles. For complex schedules or long timelines (6+ months), see resources/methodology.md for advanced scheduling techniques.

Step 3: Design retrieval practice methods

Create active recall mechanisms: flashcards, practice problems, mock tests, self-quizzing. Avoid passive techniques (highlighting, re-reading). See Common Patterns for domain-specific approaches.

Step 4: Execute daily learning sessions

Follow the schedule: new material in morning (peak alertness), reviews in afternoon/evening. Use retrieval practice consistently. Log what's difficult for extra review. For advanced techniques like interleaving or desirable difficulties, see resources/methodology.md.

Step 5: Track progress and adjust

Measure retention with self-tests. Adjust review frequency based on performance (struggle more = review sooner). Update schedule as needed. Validate using resources/evaluators/rubric_memory_retrieval_learning.json.

Common Patterns

Exam Preparation (3-6 months):

  • Phase 1 (60% time): Initial learning + comprehension
  • Phase 2 (30% time): Spaced review + retrieval practice
  • Phase 3 (10% time): Mock exams + weak area focus
  • Use: Professional certifications, academic finals, bar exam

Language Learning (ongoing):

  • Daily: 10 new vocabulary words + review old words due today
  • Weekly: Grammar lesson + interleaved practice with prior lessons
  • Monthly: Conversation practice integrating all learned material
  • Use: Spanish, Mandarin, French, any language mastery

Technology/Job Skill (3-12 weeks):

  • Week 1-2: Fundamentals + hands-on practice
  • Week 3-6: Advanced concepts + spaced review of fundamentals
  • Week 7+: Real projects + systematic review of challenging concepts
  • Use: Learning Python, React, AWS, data analysis

Medical/Technical Procedures:

  • Day 1: Learn procedure steps + immediate practice
  • Day 2: Retrieval practice without notes
  • Day 4: Practice + add edge cases
  • Day 8: Full simulation
  • Day 15, 30: Refresh to maintain
  • Use: Clinical skills, safety protocols, lab techniques

Bulk Memorization (facts, dates, lists):

  • Create spaced repetition flashcard deck
  • Review cards daily (Anki algorithm or similar)
  • Retire cards after 5+ successful recalls
  • Add mnemonic devices for difficult items
  • Use: Anatomy, geography, historical dates, pharmacology

Guardrails

Avoid Common Mistakes:

  • ❌ Passive re-reading or highlighting → Use active retrieval instead
  • ❌ Cramming (massed practice) → Use spaced repetition
  • ❌ Blocking by topic (all topic A, then all topic B) → Use interleaving
  • ❌ Over-confidence after initial learning → Test yourself repeatedly
  • ❌ No tracking → Measure retention to adjust schedule

Realistic Expectations:

  • Forgetting is normal and necessary for strong memory consolidation
  • Initial struggles with retrieval are productive ("desirable difficulties")
  • Expect 20-40% forgetting between reviews (that's the sweet spot)
  • Spaced repetition feels less productive than massing, but works better
  • Plan for 2-3x more time than you think you need

Time Management:

  • Daily consistency > marathon sessions
  • Minimum 15-20 min/day more effective than 2 hours weekly
  • Peak retention: 25 min study → 5 min break → repeat
  • Review sessions should be shorter than initial learning sessions
  • Build buffer for life interruptions (illness, travel, deadlines)

When to Seek Help:

  • Material isn't making sense after 3+ attempts → Get instructor/expert help
  • Retention remains below 60% after 3 review cycles → Reassess study method
  • Burnout or motivation collapse → Reduce daily load, add intrinsic rewards
  • Test anxiety interfering → Address anxiety separately from memory techniques

Quick Reference

Resources:

  • resources/template.md - Learning plan template with scheduling
  • resources/methodology.md - Advanced techniques for complex learning goals
  • resources/evaluators/rubric_memory_retrieval_learning.json - Quality criteria

Output:

  • File: memory-retrieval-learning.md in current directory
  • Contains: Learning goals, material breakdown, review schedule, retrieval methods, tracking system

Success Criteria:

  • Spaced repetition schedule covers entire timeline
  • Retrieval practice methods defined for all material types
  • Daily time commitment is realistic and sustainable
  • Tracking mechanism in place to measure retention
  • Schedule includes buffer for setbacks
  • Validated against quality rubric (score ≥ 3.5)

Evidence-Based Techniques:

  1. Spacing Effect: Reviews at 1, 3, 7, 14, 30 days
  2. Testing Effect: Self-test > re-study for long-term retention
  3. Interleaving: ABCABC > AAABBBCCC for transfer and discrimination
  4. Elaboration: Connect to prior knowledge, explain to others
  5. Dual Coding: Combine verbal + visual representations

平台分布

Claude Code

27.56%
按下载量换算532

Gemini CLI

20.84%
按下载量换算402

Antigravity

19.63%
按下载量换算379

windsurf

12.68%
按下载量换算245

OpenCode

8.7%
按下载量换算168

github-copilot

3.74%
按下载量换算72

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