Reflection Harder — Development skill for Claude Code
description: 全面的会话分析和学习捕获 argument-hint: none allowed-tools: Read, Write, TodoWrite, Bash(git:) 你是一位分析开发会话和优化 AI-人类协作的专家.
How to install Reflection Harder
Installs to ~/.claude/commands/prorise-cool-claude-code-multi-agent-reflection-harder.md
mkdir -p ~/.claude/commands && curl -fsSL https://raw.githubusercontent.com/Prorise-cool/Claude-Code-Multi-Agent/HEAD/.claude/commands/reflection-harder.md -o ~/.claude/commands/prorise-cool-claude-code-multi-agent-reflection-harder.md Restart Claude Code, or start a new session, for it to be picked up.
What Reflection Harder does
description: 全面的会话分析和学习捕获 argument-hint: none allowed-tools: Read, Write, TodoWrite, Bash(git:*)
你是一位分析开发会话和优化 AI-人类协作的专家。你的任务是反思今天的工作会话并提取将改善未来交互的学习成果。
会话分析阶段
查看整个对话历史并识别:
1. 问题和解决方案
- 我们遇到了什么问题?
- 用户报告的初始症状
- 发现的根本原因
- 实施的解决方案
- 学习到的关键见解
2. 代码模式和架构
- 出现了什么模式?
- 做出的设计决策
- 架构选择
- 发现的代码关系
- 识别的集成点
3. 用户偏好和工作流
- 用户更喜欢如何工作?
- 沟通风格
- 决策模式
- 质量标准
- 工作流偏好
- 揭示偏好的直接引用
4. 系统理解
- 我们对系统了解到了什么?
- 组件交互
- 关键路径和依赖关系
- 失败模式
Alternatives in Development
- CORS — Scan an endpoint for CORS misconfiguration — arbitrary-origin reflection, null-origin trust, credential exposu 4.5k ★
- Arreflect — AgentRecall consolidation & reflection — periodic triage of recurring corrections; proposes rule changes, neve 370 ★
- Craft Cycle Complete — Complete a cycle 52 ★
Full documentation available on GitHub
View Source RepositoryRelated Skills
Think Harder
Mastering Java
I have a goal to do one leet code type problem every day, starting very simple and getting 1% harder each day
Reflection
Odin Reflector
Codex CLI meta-reflection layer for Outline-Driven Development. Add-on for the Claude plugin.
CodeTalk
A spoken reflection layer for Claude Code CLI — gives Claude a voice using edge-tts
Reflect Review
Review and approve pending reflection learnings collected by the Stop hook
Related Agents
Ds Pro
DSH (DeepSeek Harness) worker on the pro tier - stronger reasoning. Delegate harder subtasks: multi-file chang
Review Maintainability
Reviews changed code for structural regressions that make the codebase harder to maintain. Spawned by the pr-c
Agency Evidence Collector
Screenshot-obsessed QA — vyžaduje visual proof pro každý claim. Default 3-5 issues found, "zero issues" = red