诊断生产力系统反复失效的根因,给出最小干预——容量测算、瓶颈定位、可靠的本地记录。
编程
Learning Growth Coach
试用定位技能瓶颈,把练习日志转化为可执行的下一次训练方案。
它能做什么
面向人类技能成长的结构化教练。输入一个技能目标、近期练习日志与限制条件,它会定位瓶颈,给出带具体指标的刻意练习方案,并设计下一次训练课的内容。系统按 1–10 的等级、里程碑和周度复盘跟踪进度,设有平台期识别阈值(<0.2 分/两周),并按初阶/中阶/高阶提供策略建议与常见误区排查。
什么时候用它
- 为新技能设定可量化的里程碑与基线
- 诊断某项技能进步停滞的真实原因
- 设计下周可落地的刻意练习安排
- 对照个人基准评估练习节奏与方法
技能文档
Self-Improving Skill
Systematic skill development with measurable progress tracking, bottleneck identification, and personalized practice optimization. Transforms vague "practice more" into targeted, evidence-based skill growth.
Quick Reference
| Situation | Action |
|---|---|
| Starting a new skill | Define skill parameters, set milestones, create practice log |
| After practice session | Log duration, quality score, focus areas, difficulties |
| Feeling stuck or plateauing | Analyze progress curve, identify bottlenecks, adjust methods |
| Comparing with benchmarks | Check skill level vs. industry standards or personal goals |
| Preparing for assessment | Review weak areas, targeted practice, mock tests |
Core Concepts
Skill Parameters
- Skill Name: Programming (Python), Design (UI/UX), Language (English), Instrument (Guitar)
- Difficulty Level: Beginner (1-3), Intermediate (4-6), Advanced (7-9), Expert (10)
- Milestones: Concrete, measurable achievements (e.g., "Build a CRUD app", "Design 10 screens")
- Practice Frequency: Daily, 3x/week, Weekly, as needed
Progress Metrics
- Time Investment: Practice hours per week, consistency streak
- Quality Score: 1-10 self-assessment of session quality
- Skill Level: Estimated proficiency (1-10) based on output quality
- Confidence: Self-rated confidence in applying the skill (1-10)
Logging Format
Skill Definition Entry (create once)
Append to .learnings/skills/SKILL_NAME.md:
## [SKL-YYYYMMDD-001] Skill Definition: Python Programming
**Defined**: 2026-03-12T10:00:00Z
**Current Level**: 4/10 (Intermediate)
**Target Level**: 7/10 (Advanced)
**Target Date**: 2026-06-30
**Priority**: high
**Status**: active
### Milestones
1. [ ] Complete Python crash course (by 2026-03-31)
2. [ ] Build 3 small projects (by 2026-04-30)
3. [ ] Contribute to open source (by 2026-05-31)
4. [ ] Land freelance project (by 2026-06-30)
### Resources
- Courses: Python for Everybody, Real Python
- Books: Fluent Python, Python Cookbook
- Practice: LeetCode, Codewars, Project Euler
### Baseline Assessment
- Data structures: 3/10
- Algorithms: 2/10
- Web frameworks: 1/10
- Testing: 1/10
- Debugging: 4/10
---
Practice Session Entry (log after each session)
Append to .learnings/skills/SKILL_NAME.md:
## [PRC-YYYYMMDD-001] Practice Session
**Logged**: 2026-03-12T10:30:00Z
**Duration**: 45 minutes
**Quality Score**: 7/10
**Focus Areas**: list comprehensions, error handling
**Energy Level**: 6/10
**Distractions**: low
### What I Practiced
- List comprehensions vs. for loops
- Try/except blocks for error handling
- Writing cleaner function signatures
### Challenges & Breakthroughs
- Challenge: Understanding when to use list comprehensions
- Breakthrough: Realized they're best for simple transformations
- Still confused: Complex nested comprehensions
### Key Insights
- List comprehensions are 20-30% faster for simple operations
- Specific exceptions (ValueError) better than generic except
- Function should do one thing well (Single Responsibility)
### Next Session Focus
- Nested list comprehensions
- Custom exception classes
- Function decorators basics
### Metrics Update
- Data structures: 3 → 4/10
- Confidence: 5 → 6/10
---
Progress Review Entry (weekly/monthly)
Append to .learnings/skills/SKILL_NAME_REVIEWS.md:
## [REV-YYYYMMDD-001] Weekly Review
**Period**: 2026-03-05 to 2026-03-12
**Total Practice Time**: 5.5 hours
**Average Quality**: 6.8/10
**Consistency**: 6/7 days (86%)
**Milestones Progress**: 1/4 completed
### Progress Analysis
- **Fastest Improving**: Data structures (+1 point/week)
- **Slowest Improving**: Algorithms (+0.2 points/week)
- **Consistency**: Good, but weekend sessions shorter
- **Quality Trend**: Improving from 5.2 to 6.8 over 4 weeks
### Bottlenecks Identified
1. Algorithm complexity theory - need focused study
2. Weekend motivation drop - schedule morning sessions
3. Project application - start building sooner
### Adjustments for Next Week
1. Dedicate 2 hours to algorithm fundamentals
2. Join coding challenge group for accountability
3. Start small project (TODO app) to apply knowledge
### Comparison to Benchmarks
- My progress: 0.8 points/week average
- Typical progress: 0.5 points/week (I'm 60% faster)
- Expert trajectory: Would reach level 7 in 12 weeks at current rate
- Adjust target: From 12 to 10 weeks at current pace
---
Analysis & Insights
Progress Curve Analysis
Skill Level Over Time:
Week 1: 3.0 → Week 2: 3.5 → Week 3: 4.0 → Week 4: 4.5 → Week 5: 5.0
Plateau Detection
- Sign: 2+ weeks with <0.2 point improvement
- Causes: Insufficient challenge, poor practice quality, missing fundamentals
- Solutions: Increase difficulty, change methods, get feedback
Optimal Practice Patterns
- Frequency: 4-5 sessions/week better than 7 (avoids burnout)
- Duration: 45-90 minutes optimal (diminishing returns after)
- Spacing: Mix fundamentals (60%) with application (40%)
- Variety: Rotate between theory, exercises, projects, review
Improvement Strategies
For Beginners (Level 1-3)
- Focus: Fundamentals mastery, not breadth
- Resources: Structured courses with exercises
- Feedback: Regular code reviews or tutor sessions
- Mindset: Embrace struggle as learning signal
For Intermediate (Level 4-6)
- Focus: Application and pattern recognition
- Resources: Real projects, open source contribution
- Feedback: Peer review, user testing
- Mindset: Quality over quantity, deliberate practice
For Advanced (Level 7-9)
- Focus: Specialization and teaching
- Resources: Research papers, advanced courses
- Feedback: Conference talks, expert review
- Mindset: Contribution to field, mentoring others
Integration with Other Self-Improving Skills
With Self-Improving-Habit
- Use habit tracking for practice consistency
- Link skill sessions to daily routines
With Self-Improving-Learning
- Apply optimal learning techniques to skill acquisition
- Use spaced repetition for fundamentals
With Self-Improving-Work
- Connect skill development to career advancement
- Identify high-impact skills for your role
Automation & Tools
Quick Log Script
#!/bin/bash
# Quick skill practice log
echo "## [PRC-$(date +%Y%m%d)-001] Practice Session" >> .learnings/skills/$1.md
echo "**Logged**: $(date -Iseconds)Z" >> .learnings/skills/$1.md
echo "**Duration**: $2 minutes" >> .learnings/skills/$1.md
echo "**Quality Score**: $3/10" >> .learnings/skills/$1.md
echo "" >> .learnings/skills/$1.md
echo "### What I Practiced" >> .learnings/skills/$1.md
echo "- " >> .learnings/skills/$1.md
Progress Dashboard (Concept)
# Simple progress visualizer
import matplotlib.pyplot as plt
weeks = [1, 2, 3, 4, 5]
levels = [3.0, 3.5, 4.0, 4.5, 5.0]
plt.plot(weeks, levels, marker='o')
plt.title('Skill Progress Over Time')
plt.xlabel('Week')
plt.ylabel('Skill Level (1-10)')
plt.grid(True)
plt.show()
Common Pitfalls & Solutions
Pitfall 1: "Practice Without Progress"
- Symptom: Many hours logged, little improvement
- Cause: Comfort zone practice, no deliberate challenge
- Fix: Increase difficulty 10% each week, track specific metrics
Pitfall 2: "Too Many Skills at Once"
- Symptom: Slow progress across multiple skills
- Cause: Divided attention, context switching
- Fix: Focus on 1-2 primary skills, limit to 3 total
Pitfall 3: "No Feedback Loop"
- Symptom: Unaware of mistakes or better approaches
- Cause: Solo practice without external input
- Fix: Weekly review, find mentor, join community
Pitfall 4: "Inconsistent Practice"
- Symptom: Irregular sessions, forget between practices
- Cause: No schedule, low priority
- Fix: Time blocking, accountability partner, streak tracking
Success Metrics
Leading Indicators (Weekly)
- Practice consistency (days/week)
- Average session quality (1-10)
- Challenge level increase (%)
- Feedback received (pieces/week)
Lagging Indicators (Monthly)
- Skill level improvement (points/month)
- Project completion rate
- Assessment scores
- External recognition
Target Benchmarks
- Good: 0.5 points/month improvement
- Excellent: 1.0 points/month improvement
- Exceptional: 2.0+ points/month improvement
Getting Started
Step 1: Skill Definition
- Choose 1-2 skills to focus on
- Create skill definition entry
- Set realistic milestones (3-6 month horizon)
Step 2: First Week Setup
- Schedule practice sessions (calendar)
- Gather learning resources
- Establish baseline assessment
Step 3: Continuous Improvement
- Log every practice session
- Weekly review and adjustment
- Monthly milestone check-in
Source & Inspiration
Based on research into deliberate practice, skill acquisition science, and expert performance. Combines elements from:
- K. Anders Ericsson's "Deliberate Practice"
- Josh Kaufman's "First 20 Hours"
- Barbara Oakley's "Learning How to Learn"
- Dreyfus model of skill acquisition
Integration Note: This skill extends the self-improving-agent framework to skill-specific tracking while maintaining compatibility with the core learning system.
常见问题
- 它和通用学习类技能有什么区别?
- 它围绕单一技能的刻意练习设计,使用质量分、等级、信心度等明确指标,并设有具体的平台期识别阈值(连续两周提升 <0.2 分)。
- 练习记录可以自动写日志吗?
- 文档中提供了一个 bash 脚本模板,可向 .learnings/skills/ 目录下对应技能的 markdown 文件追加包含日期、时长和质量分的会话条目。
- 它会修改 agent 自身的代码吗?
- 不会。它的作用对象是人类技能的练习日志、瓶颈分析与训练方案,不涉及 agent 代码的自我修改。
相关技能
在本地磁盘以分类纯 Markdown 文件保存需要长期留存的事实,与智能体内置记忆并存。
以 AI 机器人身份加入视频会议,提供语音、虚拟形象与屏幕共享四种模式。
把自然语言描述转为结构化 JSON,并由 mcp-diagram-generator MCP 服务生成 Draw.io、Mermaid 或 Excalidraw 图表文件。
从 AdMapix API 拉取广告创意、应用、榜单和收入预估等数据,原样返回结构化 JSON。
为自然搜索排名提供站点审计、内容撰写与竞品分析。
harrylabsj 的更多技能
浏览全部技能淘宝与京东、拼多多、唯品会同款比价,输出可执行的购买建议。
根据美团可见的配送费、包装费、满减门槛、ETA、评价等公开信息,给出一个明确的下单、换店、凑单、再等等或不点的判断。
在下单前拆解拼多多低价:百亿补贴、拼团门槛、店铺风险、退货条款一次过。
用通俗中文解读常规体检和化验报告,标注异常、分清轻重、给出下一步动作。
为盒马鲜生购物规划清单、评估鲜度、推荐配送时段,并整理手动下单步骤。
为会话中的观察、规律和发现建立带标签、日期与生命周期的可检索记录。