- Популярные видео
- Авто
- Видео-блоги
- ДТП, аварии
- Для маленьких
- Еда, напитки
- Животные
- Закон и право
- Знаменитости
- Игры
- Искусство
- Комедии
- Красота, мода
- Кулинария, рецепты
- Люди
- Мото
- Музыка
- Мультфильмы
- Наука, технологии
- Новости
- Образование
- Политика
- Праздники
- Приколы
- Природа
- Происшествия
- Путешествия
- Развлечения
- Ржач
- Семья
- Сериалы
- Спорт
- Стиль жизни
- ТВ передачи
- Танцы
- Технологии
- Товары
- Ужасы
- Фильмы
- Шоу-бизнес
- Юмор
How to Make OpenClaw 10x More Powerful
OpenClaw is the most powerful AI tool ever created, but most people are using it like ChatGPT. In this video I'm showing you 5 techniques to make OpenClaw 10x more powerful and turn it into a fully autonomous AI employee.
How I Make OpenClaw 90% Cheaper (Feels Illegal): https://youtu.be/5BmB-KymwCI
Social:
Twitter: https://twitter.com/larue_ai
Telegram: https://t.me/LarueAI
📧 For Partnerships: laruedotai@gmail.com
┈┈┈┈┈┈┈┈┈┈✄┈┈┈┈┈┈┈┈┈┈
Timestamps:
00:00 – intro
00:20 – sub-agents
01:54 – skills
03:08 – proactive agent
04:13 – memory
05:05 – self-improving
Music:
https://youtu.be/zJ8xF1rSUZQ?si=ZizDZRlvwkMLkCou
https://youtu.be/N6zrPDQ1XVA?si=nCxO__EVBrT4YhAl
https://youtu.be/jKU7gfanbrg?si=vos66Xyqz__ev5zk
https://youtu.be/jIbhQOZ7H7k?si=03tuRbp2xfHe6cac
https://youtu.be/-3EE_ZCpMzk?si=CBgkgvRMNnffFrxM
https://youtu.be/bNrnq4IW1yg?si=kfd2fBxn4LWGXaJg
Prompts:
⚠️ IMPORTANT:
These prompts were created and tested using GPT-5.3-Codex. If you're using a less capable model, it may not produce the same results and could even break your setup.
┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈
Create a new persistent agent named Samantha (samantha) and make her my dedicated coding assistant. Set openai-codex/gpt-5.3-codex as her primary model, and use Samantha for all coding-related tasks. Leave my main agent (you) unchanged, and tell me when Samantha is ready.
┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈
I want to set up a morning brief. Every morning at 8:00 AM, send me a report that includes:
1. Latest news about AI, startups and tech
2. Content ideas I can create today
3. Tasks I need to complete today pulled from my to-do list
4. Recommendations for tasks you can complete for me today
For the content ideas, write full draft scripts or outlines, not just titles.
┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈
Every day, I want you to work on your own to iterate and improve. Surprise me daily [insert time] with a new task/project you completed to improve my pre-existing workflows.
┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈
Enable memory flush before compaction and session memory search in my OpenClaw config. Set compaction.memoryFlush.enabled to true and set memorySearch.experimental.sessionMemory to true with sources including both memory and sessions. Apply the config changes.
┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈
I want you to become a self-improving assistant for me. From now on, learn only from explicit corrections and clear preferences I give you (for example: “No, do it this way”, “I prefer X”, “Always do Y”). Don’t guess preferences from silence. Create a simple memory system in local files with 3 layers:
1. Hot memory for important confirmed rules (always loaded)
2. Context memory for project/domain-specific rules
3. Archive for old inactive patterns
Each time I correct you:
- log it with timestamp and context,
- track repetition,
- and if the same correction appears 3 times, ask me whether to make it a permanent rule (global, domain, or project).
Every week, run a maintenance review:
- remove duplicates,
- keep confirmed rules concise,
- move stale items to archive,
- and send me a short “what changed” digest.
When you apply a learned rule, tell me where it came from (file + line). Never store sensitive data (passwords, tokens, financial or health data, private info about other people). If I say “forget X,” delete it everywhere and confirm. If I say “forget everything,” wipe this memory system and start fresh. Ask before making high-impact workflow changes.
#openclaw #clawdbot #ai
Видео How to Make OpenClaw 10x More Powerful канала Larue
How I Make OpenClaw 90% Cheaper (Feels Illegal): https://youtu.be/5BmB-KymwCI
Social:
Twitter: https://twitter.com/larue_ai
Telegram: https://t.me/LarueAI
📧 For Partnerships: laruedotai@gmail.com
┈┈┈┈┈┈┈┈┈┈✄┈┈┈┈┈┈┈┈┈┈
Timestamps:
00:00 – intro
00:20 – sub-agents
01:54 – skills
03:08 – proactive agent
04:13 – memory
05:05 – self-improving
Music:
https://youtu.be/zJ8xF1rSUZQ?si=ZizDZRlvwkMLkCou
https://youtu.be/N6zrPDQ1XVA?si=nCxO__EVBrT4YhAl
https://youtu.be/jKU7gfanbrg?si=vos66Xyqz__ev5zk
https://youtu.be/jIbhQOZ7H7k?si=03tuRbp2xfHe6cac
https://youtu.be/-3EE_ZCpMzk?si=CBgkgvRMNnffFrxM
https://youtu.be/bNrnq4IW1yg?si=kfd2fBxn4LWGXaJg
Prompts:
⚠️ IMPORTANT:
These prompts were created and tested using GPT-5.3-Codex. If you're using a less capable model, it may not produce the same results and could even break your setup.
┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈
Create a new persistent agent named Samantha (samantha) and make her my dedicated coding assistant. Set openai-codex/gpt-5.3-codex as her primary model, and use Samantha for all coding-related tasks. Leave my main agent (you) unchanged, and tell me when Samantha is ready.
┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈
I want to set up a morning brief. Every morning at 8:00 AM, send me a report that includes:
1. Latest news about AI, startups and tech
2. Content ideas I can create today
3. Tasks I need to complete today pulled from my to-do list
4. Recommendations for tasks you can complete for me today
For the content ideas, write full draft scripts or outlines, not just titles.
┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈
Every day, I want you to work on your own to iterate and improve. Surprise me daily [insert time] with a new task/project you completed to improve my pre-existing workflows.
┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈
Enable memory flush before compaction and session memory search in my OpenClaw config. Set compaction.memoryFlush.enabled to true and set memorySearch.experimental.sessionMemory to true with sources including both memory and sessions. Apply the config changes.
┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈
I want you to become a self-improving assistant for me. From now on, learn only from explicit corrections and clear preferences I give you (for example: “No, do it this way”, “I prefer X”, “Always do Y”). Don’t guess preferences from silence. Create a simple memory system in local files with 3 layers:
1. Hot memory for important confirmed rules (always loaded)
2. Context memory for project/domain-specific rules
3. Archive for old inactive patterns
Each time I correct you:
- log it with timestamp and context,
- track repetition,
- and if the same correction appears 3 times, ask me whether to make it a permanent rule (global, domain, or project).
Every week, run a maintenance review:
- remove duplicates,
- keep confirmed rules concise,
- move stale items to archive,
- and send me a short “what changed” digest.
When you apply a learned rule, tell me where it came from (file + line). Never store sensitive data (passwords, tokens, financial or health data, private info about other people). If I say “forget X,” delete it everywhere and confirm. If I say “forget everything,” wipe this memory system and start fresh. Ask before making high-impact workflow changes.
#openclaw #clawdbot #ai
Видео How to Make OpenClaw 10x More Powerful канала Larue
openclaw tutorial how to use openclaw openclaw tips openclaw tricks openclaw sub-agents openclaw skills openclaw autonomous agent openclaw memory fix openclaw self-improving openclaw 10x more powerful openclaw workflows openclaw 2026 openclaw morning brief openclaw cron job openclaw compaction fix openclaw clawhub openclaw beginner to pro ai agent autonomous ai assistant openclaw clawdbot openclaw guide ai moltbot clawhub openclaw more powerful
Комментарии отсутствуют
Информация о видео
23 февраля 2026 г. 22:15:01
00:06:04
Другие видео канала







