OpenClaw 实用建议：实现 AI Agent 的自我改进与果断决策

Skip to main content Toggle navigation menu Toggle navigation menuArticlesPodcastsVideosTweetsSourcesNewsletters

⌘K

Change language Switch ThemeSign In

Narrow Mode

OpenClaw Practical Tips: Enabling AI Agent Self-Improvement and Decisive Action ===============================================================================

OpenClaw Practical Tips: Enabling AI Agent Self-Improvement and Decisive Action ===============================================================================  ### meng shao

@shao__meng

给非开发背景的朋友们三个 OpenClaw 实用建议：让它自己修复、自己学习、自己做决定

• 实现自我改进机制

关键实践：创建 .learnings/ 文件夹，包含 ERRORS.md （记录每起故障及其上下文）和 LEARNINGS.md（记录每次修正）。在 AGENTS.md 中添加指令：“完成任何任务后，检查 .learnings/ 中相关历史错误；任何故障后，立即记录。”

这个方法模拟人类学习过程，避免 Agents 重复错误。通过文件化反馈循环，Agents 逐步积累知识，提升长期可靠性。适用于迭代任务场景，如自动化工作流，体现了 AI 的“记忆”强化策略。

• 实现 100% 崩溃修复率

关键实践：Agents 崩溃时，避免手动调试，而是向 Claude Code 或 Codex 等提供指令：“我的 OpenClaw [无响应/缓慢/出错]。SSH 进入机器，阅读日志，排查问题。行动前制定计划。”

这个技巧降低技术门槛，让非开发者也能维护复杂系统，体现了 Agents 间协作的“Meta Agent”概念，适用于生产环境下的故障管理。

• 赋予意见性风格

关键实践：编辑 SOUL.md（OpenClaw 的核心个性文件），添加规则：“采取立场，从不模棱两可（如‘你可以做X或Y’）；保持简洁，无赘述；不确定时，仅问一个澄清问题，然后执行。”

默认 OpenClaw 性格会更礼貌和听话，这个优化使其更果断、高效。类似于人类决策训练，增强 Agents 的实用性和用户体验，避免信息过载。适用于需要快速响应的应用，如咨询或自动化决策。Show More

!Image 3: Tweet image

!Image 4: Corey Ganim

#### Corey Ganim

@coreyganim · 2d ago

3 OpenClaw tips that took me months to figure out:

• Make it self-improve

Create a .learnings/ folder with:

• ERRORS.md → Log every failure with context

• LEARNINGS.md → Log every correction you make

Add this to your AGENTS.md:

"After completing ANY task, check .learnings/ for relevant past errors. After ANY failure, log it immediately."

Your agent stops making the same mistake twice.

• 100% fix rate for crashes

When OpenClaw breaks, don't debug manually. Point Claude Code at your files:

"My OpenClaw agent is [not responding / slow / erroring]. SSH into my machine, read the logs, and troubleshoot. Build a plan before taking action."

I've never written code. This works every time.

• Make it opinionated

Edit your SOUL.md:

"- Take a stance. Never hedge with "you could do X or Y"

• Be concise. No filler phrases.

• When unsure, ask one clarifying question, then execute."

Default OpenClaw is polite but vague. Fix that.

Steal these. Your agent will thank you.Show More

36

53

664

66.5K

Mar 16, 2026, 12:41 AM View on X

5 Replies

4 Retweets

30 Likes

4,442 Views  meng shao @shao__meng

One Sentence Summary

Provides three optimization tips for non-developer users of OpenClaw Agents: establishing a self-improvement mechanism, achieving automatic crash repair, and imparting a decisive personality.

Summary

This tweet offers three advanced usage tips for the AI Agent tool OpenClaw. Firstly, it suggests creating a .learnings/ folder to record errors and lessons, building AI's "long-term memory" to prevent repeated mistakes. Secondly, it recommends using tools like Claude Code for remote SSH debugging to automatically fix crashes, lowering the maintenance barrier. Finally, it advises editing SOUL.md to eliminate ambiguous Agent expressions and improve decision-making efficiency. These methods aim to enhance the Agent's reliability, autonomy, and practicality.

AI Score

83

Influence Score 16

Published At Today

Language

Chinese

Tags

OpenClaw

AI Agent

Prompt Engineering

Self-Improvement

Automated Operations HomeArticlesPodcastsVideosTweets

OpenClaw Practical Tips: Enabling AI Agent Self-Improveme... ===============