Token导航 LogoToken导航TokenDH.com
研究检索只读clawhub未标认证来源可访问clear审计通过

self-improving-cognition自我提升认知

Agent Skill

self-improving-cognition 用于记录任务执行中的错误、用户纠正、经验和能力缺口,适合在 OpenClaw 中希望让 Agent 持续沉淀问题、修正和最佳实践时使用。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。

总安装

9,180

周安装

375

GitHub Stars

公开资料未说明

下载量

2,970
OpenClaw

安装说明

本站只整理中文说明和来源信息,不托管安装包,也不代用户安装。

GitHub

来源数

2

许可证

MIT-0

最后核验

2026-05-01

来源状态

来源可访问

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:self-improving-cognition(自我提升认知)
来源仓库:https://github.com/harrylabsj/self-improving-cognition
安装命令:
openclaw skills install self-improving-cognition
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 OpenClaw 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。

ClawHubOpenClaw
openclaw skills install self-improving-cognition

简介

用于改善代理的认知模式、决策质量与问题解决能力,基于心理学原理进行优化。

  • 适用于识别认知偏差、提升判断力与增强逻辑推理的场景。
  • 通过系统化训练与反馈循环,推动代理在复杂决策中表现更优。
  • 安装命令:openclaw skills install self-improving-cognition,需确认是否依赖外部研究数据。
  • 建议结合具体任务类型定制训练目标,避免泛化失效。

SKILL.md

name
self-improving-cognition
description
Systematic improvement of thinking patterns, decision quality, problem-solving abilities, and mental performance. Use when noticing cognitive biases, making important decisions, solving complex problems, or wanting to enhance mental clarity and effectiveness.

Self-Improving Cognition

Structured approach to cognitive enhancement through measurable tracking of thinking patterns, bias identification, decision analysis, and mental skill development. Transforms vague "think better" into specific, improvable cognitive competencies.

Quick Reference

SituationAction
Making important decisionLog decision context, alternatives, reasoning, track outcome
Noticing cognitive biasIdentify bias type, record instance, develop correction strategy
Solving complex problemDocument problem space, approaches tried, breakthrough insights
Feeling mentally foggyTrack mental clarity factors, identify patterns, adjust routines
Learning new complex conceptMap understanding, identify gaps, track mastery progression

Cognitive Dimensions & Metrics

Attention & Focus (Cognitive Control)

  • Sustained Attention: Minutes of uninterrupted focus
  • Selective Attention: Ability to filter distractions (1-10)
  • Divided Attention: Task switching efficiency (1-10)
  • Focus Recovery: Time to regain focus after interruption

Memory & Recall (Information Processing)

  • Working Memory: Items held simultaneously (digit span, n-back)
  • Long-term Recall: Accuracy of factual recall (1-10)
  • Memory Speed: Time to retrieve information
  • Forgetting Curve: Retention after 1 day, 1 week

Logical Reasoning (Analytical Thinking)

  • Deductive Accuracy: Logical conclusion correctness (%)
  • Inductive Strength: Pattern recognition quality (1-10)
  • Fallacy Detection: Ability to spot reasoning errors (%)
  • Argument Quality: Structure and evidence strength (1-10)

Creative Thinking (Generative Cognition)

  • Idea Fluency: Number of ideas generated in time period
  • Idea Originality: Novelty score (1-10)
  • Concept Connection: Unusual associations made
  • Solution Elegance: Simplicity and effectiveness (1-10)

Metacognition (Thinking About Thinking)

  • Accuracy: Self-assessment vs. actual performance (% match)
  • Strategy Awareness: Understanding of own thinking processes
  • Error Detection: Ability to catch own mistakes
  • Learning Adaptation: Adjusting approach based on feedback

Logging Format

Cognitive Assessment Baseline

Append to .learnings/cognition/BASELINE.md:

## [COG-YYYYMMDD-001] Cognitive Baseline Assessment

**Assessed**: 2026-03-12T10:00:00Z
**Overall Cognitive Fitness**: 6.8/10
**Strengths**: Logical reasoning, Metacognition
**Areas for Improvement**: Sustained attention, Creative fluency

### Dimension Scores (1-10)
1. **Attention & Focus**: 5.5/10
   - Sustained: 25 minutes average
   - Selective: 6/10 (easily distracted by notifications)
   - Recovery: 3 minutes after interruption
   
2. **Memory & Recall**: 6.0/10
   - Working: 5±2 items (digit span)
   - Long-term: 7/10 (good factual recall)
   - Speed: 2.3 seconds average retrieval
   
3. **Logical Reasoning**: 8.0/10
   - Deductive: 85% accuracy
   - Inductive: 8/10 (strong pattern recognition)
   - Fallacy Detection: 75% accuracy
   
4. **Creative Thinking**: 5.0/10
   - Fluency: 12 ideas/10 minutes
   - Originality: 4/10 (mostly conventional)
   - Connections: 3 unusual associations/ session
   
5. **Metacognition**: 7.5/10
   - Accuracy: 80% (self-assessment vs. performance)
   - Strategy Awareness: 8/10 (knows thinking processes)
   - Error Detection: 7/10 (catches 70% of own errors)

### Cognitive Bias Inventory
- **Confirmation Bias**: Moderate (seeks confirming evidence)
- **Anchoring**: High (influenced by initial information)
- **Availability Heuristic**: Moderate (overweights recent examples)
- **Planning Fallacy**: High (underestimates time requirements)
- **Sunk Cost Fallacy**: Low (willing to cut losses)

### Environmental & Lifestyle Factors
- **Sleep**: 6.5/10 quality, 7 hours average
- **Nutrition**: 7/10 (balanced, but afternoon sugar)
- **Exercise**: 5/10 (3x/week, inconsistent)
- **Stress**: 6/10 (moderate work pressure)
- **Mental Stimulation**: 7/10 (varied but not challenging)

### Improvement Priorities
1. **Primary**: Increase sustained attention to 40+ minutes
2. **Secondary**: Boost creative fluency to 20+ ideas/10min
3. **Tertiary**: Reduce planning fallacy impact
4. **Supporting**: Improve sleep quality to 8/10

---

Decision Analysis Log

Append to .learnings/cognition/DECISIONS.md:

## [DEC-YYYYMMDD-001] Career Path Decision

**Decision Date**: 2026-03-12
**Importance**: High (career direction)
**Timeframe**: 6-12 month impact
**Status**: Decided (Option B)

### Context & Problem
Choose between:
- **Option A**: Stay current role (stable, limited growth)
- **Option B**: Take promotion (more stress, faster growth)
- **Option C**: Switch companies (unknown, potentially higher pay)

### Decision Process
1. **Information Gathering** (3 days)
   - Talked to 5 people in similar positions
   - Researched market salaries
   - Assessed personal tolerance for stress
   
2. **Criteria Weighting**
   - Growth potential: 30%
   - Work-life balance: 25%
   - Compensation: 20%
   - Learning opportunity: 15%
   - Team quality: 10%
   
3. **Option Scoring** (1-10 weighted)
   - Option A: 6.2 (stable but stagnant)
   - Option B: 7.8 (growth, manageable stress)
   - Option C: 6.5 (high risk, high potential reward)
   
4. **Cognitive Bias Check**
   - Status Quo Bias: Considered (Option A attractive due to comfort)
   - Loss Aversion: Addressed (willing to risk for growth)
   - Overconfidence: Guarded against (consulted others)
   - Sunk Cost: Irrelevant (no prior investment)

### Reasoning & Rationale
- **Option B chosen** because:
  1. Aligns with 3-year career goal (management track)
  2. Stress increase tolerable (15% vs. 50% more pay)
  3. Company investment in my growth (training budget)
  4. Backup plan exists (can return to individual contributor if needed)

### Confidence Level & Uncertainty
- **Confidence**: 8/10 in decision quality
- **Key Uncertainties**: 
  - Actual stress level in new role (estimated)
  - Team dynamics with new reports (unknown)
  - Company stability (market conditions)
- **Contingency Plans**:
  - 3-month review: If stress > 7/10, implement coping strategies
  - 6-month exit: If miserable, start job search with new title

### Expected Outcome vs. Actual Tracking
- **Expected**: 20% skill growth, 50% pay increase, stress +2 points
- **Actual**: [To be filled after 3 months]
- **Variance Analysis**: [To be filled]

### Cognitive Lessons
1. **Bias Management**: Successfully resisted status quo bias
2. **Decision Quality**: Structured approach improved confidence
3. **Information Sufficiency**: Gathered enough but not too much
4. **Emotion Integration**: Acknowledged fear but didn't let it decide

---

Problem-Solving Log

Append to .learnings/cognition/PROBLEMS.md:

## [PROB-YYYYMMDD-001] System Architecture Design Problem

**Problem Date**: 2026-03-12
**Complexity**: High (multiple constraints, novel requirements)
**Time Spent**: 8 hours over 2 days
**Solution Status**: Resolved (hybrid approach)

### Problem Definition
Design scalable notification system supporting:
- 1M+ users
- 10+ notification types
- <100ms latency requirement
- 99.99% reliability
- Cost < $500/month at scale

### Solution Approaches Considered
1. **Approach A**: Single monolithic service
   - Pros: Simple, fast to build
   - Cons: Hard to scale, single point of failure
   - Viability: Low (scaling constraints)
   
2. **Approach B**: Microservices architecture
   - Pros: Scalable, fault-tolerant
   - Cons: Complex, operational overhead
   - Viability: Medium (over-engineered for MVP)
   
3. **Approach C**: Serverless functions + queue
   - Pros: Cost-effective, auto-scaling
   - Cons: Cold start latency, vendor lock-in
   - Viability: High (meets requirements)

4. **Approach D**: Hybrid (Approach C + simple service)
   - Pros: Balances scalability and simplicity
   - Cons: More moving parts
   - Viability: Selected

### Breakthrough Insights
1. **Core Realization**: 80% of notifications are 3 types → optimize those
2. **Pattern Recognition**: Similar to email sending systems (existing patterns)
3. **Constraint Reframing**: Reliability requirement actually means "eventual consistency with retries"
4. **Simplification**: Batch processing acceptable for non-urgent notifications

### Solution Architecture
- **Urgent notifications** (<100ms): Dedicated service (5% of traffic)
- **Standard notifications**: Serverless functions + queue (95% of traffic)
- **Fallback**: Retry queue with exponential backoff
- **Monitoring**: Latency and error tracking for each type

### Cognitive Process Analysis
- **Initial Impasse**: Overwhelmed by constraints (30 minutes stuck)
- **Breakthrough Method**: Whiteboard diagramming → pattern emergence
- **Critical Thinking**: Questioned each requirement's validity
- **Creative Synthesis**: Combined approaches rather than choosing one

### Performance Metrics
- **Design Quality**: 8/10 (meets all requirements, elegant compromise)
- **Cognitive Effort**: High (complex trade-off analysis)
- **Time Efficiency**: Good (8 hours for complex problem)
- **Solution Novelty**: Medium (adapted existing patterns)

### Improvement Opportunities
1. **Faster Pattern Recognition**: Study more system design patterns
2. **Better Constraint Analysis**: Formalize requirement prioritization
3. **Reduced Overthinking**: Set timebox for decision making
4. **Collaborative Thinking**: Involve others earlier for perspective

---

Bias Identification & Correction

Append to .learnings/cognition/BIASES.md:

## [BIAS-YYYYMMDD-001] Confirmation Bias in Project Estimation

**Identified**: 2026-03-12T14:30:00Z
**Bias Type**: Confirmation Bias
**Context**: Software project timeline estimation
**Impact**: Moderate (2-week underestimation)

### Bias Manifestation
- **Situation**: Estimating new feature development time
- **Behavior**: Sought evidence supporting optimistic timeline
- **Ignored**: Previous similar projects that took longer
- **Result**: 4-week estimate (actual likely 6 weeks)

### Detection Process
1. **Trigger**: Felt too confident about estimate
2. **Check**: Asked "What evidence contradicts this?"
3. **Discovery**: Found 3 similar past projects averaged 6 weeks
4. **Recognition**: Realized seeking confirming evidence only

### Correction Applied
1. **Forced Consideration**: Listed all past similar projects
2. **Outsider View**: Asked "What would I estimate for someone else?"
3. **Pre-mortem**: Imagined project failed, identified causes
4. **Adjustment**: Revised to 6 weeks with risk buffer

### Root Cause Analysis
- **Motivation**: Wanted to please stakeholders with fast timeline
- **Pattern**: Recurring in estimation situations (3rd instance this year)
- **Environment**: Pressure to deliver quickly, reward for optimism
- **Cognitive Style**: Prefers action over caution

### Prevention Strategy
1. **Checklist**: Always review past similar projects before estimating
2. **Devil's Advocate**: Assign someone to challenge optimistic estimates
3. **Buffer Rule**: Add 30% to initial optimistic estimate
4. **Calendar Review**: Monthly review of estimates vs. actuals

### Bias Strength Assessment
- **Before Correction**: 8/10 (strong bias influence)
- **After Correction**: 3/10 (minimal residual influence)
- **Improvement**: 5-point reduction through structured process
- **Durability**: Expected to last 2-3 months before reinforcement needed

### Related Biases to Monitor
- **Planning Fallacy**: Often co-occurs with confirmation bias
- **Optimism Bias**: Similar root in underestimating difficulties
- **Anchoring**: Initial estimate anchors subsequent thinking

---

Cognitive Training Framework

Attention Training Protocol

  • Pomodoro Technique: 25min focus, 5min break (measure interruptions)
  • Deep Work Blocks: 90min uninterrupted sessions (track quality)
  • Distraction Management: Phone away, website blockers (measure effectiveness)
  • Mindfulness Practice: 10min daily meditation (track focus improvements)

Memory Enhancement System

  • Spaced Repetition: Anki or similar for factual recall
  • Method of Loci: Memory palace for complex information
  • Chunking Practice: Grouping information into meaningful units
  • Recall Testing: Regular retrieval practice vs. re-reading

Logical Reasoning Exercises

  • Logic Puzzles: Daily puzzle solving (track speed/accuracy)
  • Argument Analysis: Critique articles/arguments (identify fallacies)
  • Deduction Games: Games like Mastermind (pattern deduction)
  • Scientific Thinking: Formulate and test hypotheses

Creativity Development

  • Idea Generation: Daily 10-minute brainstorming (count ideas)
  • Alternative Uses Test: Find novel uses for common objects
  • Concept Combination: Merge unrelated concepts (track originality)
  • Constraint Removal: "What if X limitation didn't exist?" exercises

Metacognition Building

  • Thinking Journals: Document thought processes for decisions
  • Error Logs: Record and categorize thinking errors
  • Prediction Calibration: Make predictions, compare to outcomes
  • Strategy Evaluation: Assess effectiveness of thinking approaches

Cognitive Performance Factors

Lifestyle Optimization

  • Sleep: 7-9 hours, consistent schedule (track cognitive impact)
  • Nutrition: Balanced, regular meals, hydration (energy levels)
  • Exercise: 150min/week moderate (cognitive benefits)
  • Stress Management: Techniques, boundaries (prevent impairment)

Environmental Design

  • Workspace: Organized, minimal distractions (focus support)
  • Tools: Appropriate for cognitive task (reduce mental load)
  • Timing: Match tasks to circadian rhythms (productivity alignment)
  • Social: Collaboration vs. solo based on task needs

Cognitive Load Management

  • Chunking: Break complex tasks into manageable units
  • Externalization: Write down to free working memory
  • Automation: Develop routines for repetitive decisions
  • Delegation: Offload appropriate cognitive tasks

Integration with Other Skills

With Self-Improving-Learning

  • Apply optimal learning techniques based on cognitive science
  • Match learning methods to cognitive strengths
  • Use metacognition to optimize learning strategies

With Self-Improving-Work

  • Apply clear thinking to work decisions and problem-solving
  • Use cognitive principles for productivity and focus
  • Match work tasks to cognitive energy levels

With Self-Improving-Habit

  • Build habits supporting cognitive performance (sleep, exercise)
  • Use habit stacking for cognitive training routines
  • Reduce decision fatigue through habit automation

Success Metrics

Performance Metrics

  • Attention Span: Sustained focus time (increase target)
  • Memory Accuracy: Recall correctness (% improvement)
  • Decision Quality: Outcome vs. expectation (closer alignment)
  • Problem-Solving Speed: Time to effective solution (reduction)
  • Bias Detection: Frequency and correction rate (increase)

Process Metrics

  • Metacognitive Accuracy: Self-assessment vs. reality (% match)
  • Cognitive Strategy Use: Variety and appropriateness
  • Bias Awareness: Identification frequency and speed
  • Adaptation Rate: Speed of adjusting thinking approaches

Outcome Metrics

  • Better Decisions: Improved life/work outcomes
  • Reduced Errors: Fewer mistakes from poor thinking
  • Increased Innovation: More novel solutions generated
  • Enhanced Learning: Faster skill/knowledge acquisition

Getting Started

Step 1: Cognitive Baseline (Week 1)

  1. Complete baseline assessment across 5 dimensions
  2. Identify top 2 cognitive strengths and weaknesses
  3. Select 1-2 priority areas for improvement
  4. Establish initial metrics and tracking system

Step 2: Targeted Training (Weeks 2-4)

  1. Implement specific exercises for priority areas
  2. Daily tracking of cognitive performance metrics
  3. Weekly review of progress and adjustments
  4. Environmental optimization for cognitive support

Step 3: Integration & Maintenance (Month 2+)

  1. Incorporate cognitive techniques into daily work/life
  2. Regular bias checks in important decisions
  3. Quarterly reassessment of cognitive fitness
  4. Continuous refinement of thinking strategies

Scientific Foundation

Based on:

  • Dual Process Theory (Kahneman): System 1 vs. System 2 thinking
  • Cognitive Bias Research (Tversky & Kahneman): Heuristics and biases
  • Working Memory Model (Baddeley): Cognitive architecture
  • Deliberate Practice (Ericsson): Expertise development
  • Metacognition Research (Flavell): Thinking about thinking

Integration Note: This skill provides the thinking toolkit for other self-improving skills, ensuring decisions and problem-solving are based on clear, bias-aware cognition rather than intuition or habit alone.

适合场景

01

OpenClaw 用户查找和安装 Skill 时

02

用户想查找某类 Agent Skill 时

03

需要根据任务场景推荐可安装能力包时

04

需要对比不同来源的安装命令和来源信息时

能力概览

能力 1

按任务关键词查找相关 Skills

能力 2

展示可复制的安装命令

能力 3

保留来源站点、仓库和原始说明,方便继续核验

能力 4

补充不同宿主或平台的使用分布数据

能力 5

展示第三方安全扫描或审计结果

安装后应在对应宿主中按原始 README 的触发条件使用;具体调用方式请以来源页面和 README 为准。

平台分布

OpenClaw

82.37%
按下载量换算2,446

安全审计

VirusTotal

通过

ClawScan

通过

Static analysis

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。当前只有一个来源,正式发布前建议补源仓库或其他目录站核验。

来源信息

继续浏览同类 Skills