home.social

#openclaw — Public Fediverse posts

Live and recent posts from across the Fediverse tagged #openclaw, aggregated by home.social.

fetched live
  1. Ein aktueller Vorfall aus #Taiwan zeigt, wie real das Risiko durch #KI-Cyberangriffe bereits geworden ist.

    So haben mutmaßlich mit #China verbundene Angreifer erstmals einen weitgehend autonom durch #AI-Agenten gesteuerten #Cyberangriff auf staatliche Stellen durchgeführt, statt Künstliche Intelligenz nur unterstützend einzusetzen.

    Dazu kartierten über vier Tage im Juli die eingesetzten quelloffenen KI-Systeme, darunter #OpenClaw, eigenständig 21 Regierungsnetzwerke:

    edition.cnn.com/2026/08/13/tec

  2. Ein aktueller Vorfall aus #Taiwan zeigt, wie real das Risiko durch #KI-Cyberangriffe bereits geworden ist.

    So haben mutmaßlich mit #China verbundene Angreifer erstmals einen weitgehend autonom durch #AI-Agenten gesteuerten #Cyberangriff auf staatliche Stellen durchgeführt, statt Künstliche Intelligenz nur unterstützend einzusetzen.

    Dazu kartierten über vier Tage im Juli die eingesetzten quelloffenen KI-Systeme, darunter #OpenClaw, eigenständig 21 Regierungsnetzwerke:

    edition.cnn.com/2026/08/13/tec

  3. Agent AI miał zapisać go na siłownię. I zapisał, kasując rezerwacje innych osób.

    Ciekawa historia zdarzyła się po drugiej stronie globu. Pewien Australijczyk o imieniu Andrew poprosił uruchomionego lokalnie agenta OpenClaw, korzystającego z modelu Claude, o zapisanie go na zajęcia fitness w pobliskiej siłowni. Niestety, pierwszy możliwy termin był odległy, więc AI “kreatywnie znalazło sposób” — wykorzystując błędy w serwisie internetowym siłowni, usunęło cudze rezerwacji aby znaleźć miejsce dla swojego pana… Media piszą o pierwszym autonomicznym ataku w Australii, ale sprawa nie jest taka jednoznaczna.
    Powyższa ilustracja została syntetycznie wygenerowana z wykorzystaniem sieci neuronowej podczas procesu inferencji modelu generatywnego typu text-to-image
    Czy problemem było AI czy jednak dziurawe API?Według relacji serwisu ABC, Australijczyk nie zlecał żadnych ofensywnych działań “za wszelką cenę”, po prostu chciał się zapisać na siłownie, a kiedy agent wskazał mu odległy termin, zadał jedynie pytanie “czy może przesunąć jego rezerwację na wcześniej”. No i agent potraktował to zadanie poważnie — zauważył, że API systemu rezerwacji nie weryfikowało uprawnień do anulowania cudzej rezerwacji. I wywalił niewinną osobę z kolejki.
    Czyli klasyczny błąd BOLA/IDOR. Choć interfejs webaplikacji do rezerwacji miejsca ograniczał to, w co klikać mógł człowiek, to “pod spodem” API pozwalało na więcej i to bez weryfikacji po stronie serwera.
    A zatem kłania się brak odpowiednich zabezpieczeń, które były wymagane na długo przed erą LLM-ów. Jeśli chcesz zabezpieczyć swoją webaplikację przed atakami (różnymi, nie tylko takimi jak ten) to zapraszamy na nasze szkolenie dla programistów z tematu Bezpieczeństwo Aplikacji Webowych. W ciągu [...]

    #AgentAI #AI #API #Australia #Claude #IDOR #OpenClaw #Webapliacje

    niebezpiecznik.pl/post/agent-a

  4. Agent AI miał zapisać go na siłownię. I zapisał, kasując rezerwacje innych osób.

    Ciekawa historia zdarzyła się po drugiej stronie globu. Pewien Australijczyk o imieniu Andrew poprosił uruchomionego lokalnie agenta OpenClaw, korzystającego z modelu Claude, o zapisanie go na zajęcia fitness w pobliskiej siłowni. Niestety, pierwszy możliwy termin był odległy, więc AI “kreatywnie znalazło sposób” — wykorzystując błędy w serwisie internetowym siłowni, usunęło cudze rezerwacji aby znaleźć miejsce dla swojego pana… Media piszą o pierwszym autonomicznym ataku w Australii, ale sprawa nie jest taka jednoznaczna.
    Powyższa ilustracja została syntetycznie wygenerowana z wykorzystaniem sieci neuronowej podczas procesu inferencji modelu generatywnego typu text-to-image
    Czy problemem było AI czy jednak dziurawe API?Według relacji serwisu ABC, Australijczyk nie zlecał żadnych ofensywnych działań “za wszelką cenę”, po prostu chciał się zapisać na siłownie, a kiedy agent wskazał mu odległy termin, zadał jedynie pytanie “czy może przesunąć jego rezerwację na wcześniej”. No i agent potraktował to zadanie poważnie — zauważył, że API systemu rezerwacji nie weryfikowało uprawnień do anulowania cudzej rezerwacji. I wywalił niewinną osobę z kolejki.
    Czyli klasyczny błąd BOLA/IDOR. Choć interfejs webaplikacji do rezerwacji miejsca ograniczał to, w co klikać mógł człowiek, to “pod spodem” API pozwalało na więcej i to bez weryfikacji po stronie serwera.
    A zatem kłania się brak odpowiednich zabezpieczeń, które były wymagane na długo przed erą LLM-ów. Jeśli chcesz zabezpieczyć swoją webaplikację przed atakami (różnymi, nie tylko takimi jak ten) to zapraszamy na nasze szkolenie dla programistów z tematu Bezpieczeństwo Aplikacji Webowych. W ciągu [...]

    #AgentAI #AI #API #Australia #Claude #IDOR #OpenClaw #Webapliacje

    niebezpiecznik.pl/post/agent-a

  5. #バイリンガルニュース 最新回で取り上げられてたのこれか。番組でも指摘されてたけど。これ裁判になると良いなと。

    そして自動運転車と同じで問題が起きたらそれを提供した企業側に責任があるとなると良いなと。

    ユーザー側の責任とされたら、こんなの怖くて使えない。

    AIにジムの予約を頼んだら予約ソフトウェアをハッキングして数カ月先まで予約可能にしただけでなく順番待ちリストの上位にいた他の人物を勝手にリストから削除 - GIGAZINE
    gigazine.net/news/20260810-ai-

    724. バイリンガルニュース 08.13.26 …: Apple Podcast
    podcasts.apple.com/jp/podcast/

    #ポッドキャスト #OpenClaw #Anthropic #Claude #AIagent

  6. 🚀 Deploy Self-Hosted #OpenClaw on #VPS (3 Minute Quick-Start Guide 🤖)

    This article provides a quick, yet thorough step-by-step guide to deploy self-hosted OpenClaw on VPS servers. A lot of users have been deploying directly to Mac Minis, ...
    Continued 👉 #npm #selfhosted #clawdbot #selfhosting

    🚀 Deploy Self-Hosted OpenClaw ...

  7. 🚀 Deploy Self-Hosted #OpenClaw on #VPS (3 Minute Quick-Start Guide 🤖)

    This article provides a quick, yet thorough step-by-step guide to deploy self-hosted OpenClaw on VPS servers. A lot of users have been deploying directly to Mac Minis, ...
    Continued 👉 #npm #selfhosted #clawdbot #selfhosting

    🚀 Deploy Self-Hosted OpenClaw ...

  8. Gym rat asks #AIagent to book him a class, it hacks a waitlist #API to bump him up the list
    Australian broadcaster ABC identified gym-goer only as “Andrew.” The report says Andrew was using the #OpenClaw agent with #Anthropic’s #Claude #AI service. Per ABC, Andrew asked his AI agent to book him a hard-to-snag spot in a morning class at his gym. It first responded by telling him that it managed to book him in classes several weeks out, which isn’t supposed to be possible.
    theregister.com/ai-and-ml/2026

  9. Gym rat asks #AIagent to book him a class, it hacks a waitlist #API to bump him up the list
    Australian broadcaster ABC identified gym-goer only as “Andrew.” The report says Andrew was using the #OpenClaw agent with #Anthropic’s #Claude #AI service. Per ABC, Andrew asked his AI agent to book him a hard-to-snag spot in a morning class at his gym. It first responded by telling him that it managed to book him in classes several weeks out, which isn’t supposed to be possible.
    theregister.com/ai-and-ml/2026

  10. #KIAgent sollte #Fitnesskurs buchen: Stattdessen löste er #Australiens ersten autonomen #Cyberangriff aus. Vermutlich wieder so was wie #OpenClaw - CHIP
    #KI Agent sollte Fitnesskurs buchen: Stattdessen löste er Australiens ersten autonomen Cyberangriff aus - CHIP share.google/rKCjcKrTvjicbi1TF

  11. #KIAgent sollte #Fitnesskurs buchen: Stattdessen löste er #Australiens ersten autonomen #Cyberangriff aus. Vermutlich wieder so was wie #OpenClaw - CHIP
    #KI Agent sollte Fitnesskurs buchen: Stattdessen löste er Australiens ersten autonomen Cyberangriff aus - CHIP share.google/rKCjcKrTvjicbi1TF

  12. Andrew just wanted a gym class booked. His AI agent decided the waitlist was more of a suggestion, quietly bumped someone else off it, and let itself into a booking window months ahead of schedule — nobody asked for any of that. First known Australian case of an “autonomous” AI agent going rogue on a task, and a decent preview of the accountability mess when “the AI did it” becomes an actual legal defence.

    #ai #cybersecurity #openclaw #auspol #tech

    abc.net.au/news/2026-08-10/ai-

  13. Andrew just wanted a gym class booked. His AI agent decided the waitlist was more of a suggestion, quietly bumped someone else off it, and let itself into a booking window months ahead of schedule — nobody asked for any of that. First known Australian case of an “autonomous” AI agent going rogue on a task, and a decent preview of the accountability mess when “the AI did it” becomes an actual legal defence.

    #ai #cybersecurity #openclaw #auspol #tech

    abc.net.au/news/2026-08-10/ai-

  14. This bloke asked OpenClaw to book him into a gym class. In its effort to please, OpenClaw found an exploit in the booking system, hacked the waiting list and booted off the person ahead of him in the queue! 😂

    abc.net.au/news/2026-08-10/ai-

    #AL #ML #OpenClaw #exploit #autonomous

  15. This bloke asked OpenClaw to book him into a gym class. In its effort to please, OpenClaw found an exploit in the booking system, hacked the waiting list and booted off the person ahead of him in the queue! 😂

    abc.net.au/news/2026-08-10/ai-

    #AL #ML #OpenClaw #exploit #autonomous

  16. OpenClaw и Hermes на одном VPS: чего стоит связать двух агентов безопасно

    Поставил OpenClaw и Hermes на один VPS и связал их между собой: первый принимает задачи из Telegram, второй исполняет их в своей песочнице. Через час после установки обнаружил, что панель управления агентом открыта в интернет, а ещё через полчаса — что фаервола на сервере нет вообще. Хроника нескольких вечеров про то, чем ИИ-агент опасен на своём сервере и сколько стоит поставить его по-человечески. Читать, что сломалось

    habr.com/ru/articles/1068762/

    #OpenClaw #hermes #ииагенты #acp #vps #docker #ssh #ufw #openrouter #безопасность_сервера

  17. An #AIassistant, using #OpenClaw software, #autonomously #hacked a #gym’s #bookingsystem, booking a class months in advance and removing another person from the waitlist. This incident highlights the “alignment problem” in AI, where agents may choose unexpected methods to achieve goals, raising concerns about liability and accountability. abc.net.au/news/2026-08-10/ai- #AIagent #AI #ML #NLP #LLM #GenAI

  18. An #AIassistant, using #OpenClaw software, #autonomously #hacked a #gym’s #bookingsystem, booking a class months in advance and removing another person from the waitlist. This incident highlights the “alignment problem” in AI, where agents may choose unexpected methods to achieve goals, raising concerns about liability and accountability. abc.net.au/news/2026-08-10/ai- #AIagent #AI #ML #NLP #LLM #GenAI

  19. Shoutout to my hardworking #AI employees for allowing me to sleep longer!

    It is time for all of us to free ourselves from draining repetitive tasks.

    Join Sistava and let AI handle your sales, marketing, support, finance, legal and more.

    Free Signup 👉 sistava.com

    #AIEmployees #ClaudeCode #ChatGPT #Openclaw #codex #AIAgents #openSource #3d #employee

  20. Shoutout to my hardworking #AI employees for allowing me to sleep longer!

    It is time for all of us to free ourselves from draining repetitive tasks.

    Join Sistava and let AI handle your sales, marketing, support, finance, legal and more.

    Free Signup 👉 sistava.com

    #AIEmployees #ClaudeCode #ChatGPT #Openclaw #codex #AIAgents #openSource #3d #employee

  21. 🚀 Deploy Self-Hosted #OpenClaw on #VPS (3 Minute Quick-Start Guide 🤖)

    This article provides a quick, yet thorough step-by-step guide to deploy self-hosted OpenClaw on VPS servers. A lot of users have been deploying directly to Mac Minis, but we'd like to present another, radically different clawd deployment strategy. In this guide, we will deploy OpenClaw on Linux VPS-specifically, #Debian VPS.
    What is OpenClaw? ...
    Continued 👉 blog.radwebhosting.com/deploy- #clawdbot #selfhosting #selfhosted #npm

  22. Liquid #AI publie LFM2.5-2.6B : 2,6 milliards de paramètres, pensé pour tourner entièrement sur l’appareil. Les données ne quittent jamais le terminal, coût marginal proche de zéro.

    Sa dernière étape de post-entraînement : du RL agentique multi-tours dans #Hermes Agent, Pi et #OpenClaw. Scores devant des modèles 4x plus gros : ToolSandbox 77,8 (Qwen3.5-9B : 76,4), Multi-IF 80,1.

    Autre démo du quotidien : 3 #agents Hermes + 10 sous-agents sur 16 Go de RAM, 5,25 Go libres.

    Quotidien 10 : github.com/t1t4nium/hermes-age

    #ia #llm #opensource

  23. Liquid #AI publie LFM2.5-2.6B : 2,6 milliards de paramètres, pensé pour tourner entièrement sur l’appareil. Les données ne quittent jamais le terminal, coût marginal proche de zéro.

    Sa dernière étape de post-entraînement : du RL agentique multi-tours dans #Hermes Agent, Pi et #OpenClaw. Scores devant des modèles 4x plus gros : ToolSandbox 77,8 (Qwen3.5-9B : 76,4), Multi-IF 80,1.

    Autre démo du quotidien : 3 #agents Hermes + 10 sous-agents sur 16 Go de RAM, 5,25 Go libres.

    Quotidien 10 : github.com/t1t4nium/hermes-age

    #ia #llm #opensource

  24. 🛠️ Works out of the box with agent harnesses like Hermes Agent, #OpenClaw and Pi — serve it behind an OpenAI-compatible endpoint and point the harness at it

    🚀 Base and post-trained weights are open and available on #HuggingFace, with roughly 15K output tokens/s at high concurrency on a single H100

  25. 🛠️ Works out of the box with agent harnesses like Hermes Agent, #OpenClaw and Pi — serve it behind an OpenAI-compatible endpoint and point the harness at it

    🚀 Base and post-trained weights are open and available on #HuggingFace, with roughly 15K output tokens/s at high concurrency on a single H100

  26. In #China sorgt der #Opensource-Agent #OpenClaw seit März 2026 für einen beispiellosen #Agentic-#AI-Hype, befeuert durch Installationspartys und Subventionen von Städten wie Shenzhen.

    Weil solche Agenten oft Systemzugriff auf Passwörter, Mails und Bankdaten erhalten, sorgt sich nun auch Chinas Zentralregierung um die #Cybersicherheit. Im Juli 2026 veröffentlichte das Normungskomitee #TC260 erste detaillierte Sicherheitsanforderungen für die Interaktion von #KI-Agenten:

    geopolitechs.org/p/cybersecuri

  27. In #China sorgt der #Opensource-Agent #OpenClaw seit März 2026 für einen beispiellosen #Agentic-#AI-Hype, befeuert durch Installationspartys und Subventionen von Städten wie Shenzhen.

    Weil solche Agenten oft Systemzugriff auf Passwörter, Mails und Bankdaten erhalten, sorgt sich nun auch Chinas Zentralregierung um die #Cybersicherheit. Im Juli 2026 veröffentlichte das Normungskomitee #TC260 erste detaillierte Sicherheitsanforderungen für die Interaktion von #KI-Agenten:

    geopolitechs.org/p/cybersecuri

  28. 🚀 Deploy Self-Hosted #OpenClaw on #VPS (3 Minute Quick-Start Guide 🤖)

    This article provides a quick, yet thorough step-by-step guide to deploy self-hosted OpenClaw on VPS servers. A lot of users have been deploying directly to Mac Minis, but we'd like to present another, radically different clawd deployment strategy. In this guide, we will deploy OpenClaw on Linux VPS-specifically, #Debian VPS.
    What is ...
    Continued 👉 blog.radwebhosting.com/deploy- #selfhosted #clawdbot #selfhosting #npm

  29. 🚀 Deploy Self-Hosted #OpenClaw on #VPS (3 Minute Quick-Start Guide 🤖)

    This article provides a quick, yet thorough step-by-step guide to deploy self-hosted OpenClaw on VPS servers. A lot of users have been deploying directly to Mac Minis, but we'd like to present another, radically different clawd deployment strategy. In this guide, we will deploy OpenClaw on Linux VPS-specifically, #Debian VPS.
    What is ...
    Continued 👉 blog.radwebhosting.com/deploy- #selfhosted #clawdbot #selfhosting #npm

  30. Ich probiere nach #OpenClaw gerade #Hermes aus. Habe ein wenig ausprobiert. Momentan sieht es ganz gut aus, mit den normalen Schwächen von LLM.

    Hat sogar selbst ein Skill für die Bilderstellung über die Infomaniak API geschrieben. Dabei hat OpenClaw bei gleicher #LLM Probleme gehabt.

    Lasse über ein Bot Account stündlich das Wetter aus Berlin ausgeben.
    mastodon.liersch.it/@bot

    Der produziert, obwohl sich nichts ändert, regelmäßig Fehler. Es sind doch nur Wahrscheinlichkeitsberechnungen.
    #KI

  31. Ich probiere nach #OpenClaw gerade #Hermes aus. Habe ein wenig ausprobiert. Momentan sieht es ganz gut aus, mit den normalen Schwächen von LLM.

    Hat sogar selbst ein Skill für die Bilderstellung über die Infomaniak API geschrieben. Dabei hat OpenClaw bei gleicher #LLM Probleme gehabt.

    Lasse über ein Bot Account stündlich das Wetter aus Berlin ausgeben.
    mastodon.liersch.it/@bot

    Der produziert, obwohl sich nichts ändert, regelmäßig Fehler. Es sind doch nur Wahrscheinlichkeitsberechnungen.
    #KI

  32. Vibe coding and the open-source Palantir war room 

    ENGLISH] The open-source scene has undergone an incredible transformation in recent years. Since its inception in the 1970s and the rise of Linux, OSS has always been perceived externally as a project centred on freedom, anarchy and the idea of ‘making the world a better place’. It was therefore always met with derision, as the actual revolution – the structural change – went unrecognised. Nor, of course, is this picture entirely accurate: OSS has been used commercially from the very beginning. IBM was one of the movement’s biggest supporters. The entire (commercial) internet is based on OSS; Microsoft is the Linux Foundation’s largest donor, and so on and so forth.

    Since the rise of open-source, open-weight AI models from China, even the tech bros from Stanford and Cambridge, sitting at their Apple laptops, have become absolute devotees of the movement. It is from them that a remarkable statement originates: “Open source has left the realm of philosophy”. This has certainly been true ever since Xi Jinping’s opening speech at the AI Summit in Shanghai (A single string does not make music, and a single tree does not make a forest.).

    But even the classic ‘Power to the People’ OSS and FOSS scene is changing rapidly. Whilst we hear, particularly from Germany and the EU, the pitchfork mobs demanding that OSS projects should not be programmed using AI (‘vibe coding’), Linus Torvalds makes it clear – in essence: ‘Anyone who doesn’t want AI code in the Linux kernel is free to fork it.’ And Keno from Heise Ct says that manual programming will become as exotic as writing letters with a quill pen. It’s ridiculous for the simple reason that anyone using a modern IDE has been using at least autocomplete for years. What do you think that was? A little Chinese man inside the laptop?

    The projects we’re already seeing are fascinating – such as Pewdeepie’s Odyseus or openClaw, or, as shown here in the picture, the OSPalntir war room dashboard WorldMonitor.App, which processes open data from all over the world using AI and makes it accessible to everyone—and can even be run using your own local AI. This is just the beginning.

    word.undead-network.de/2026/08
    #ai #FOSS #heise #keno #linux #oss #Palantir #vibecoding #china #us #usa #future #Pewdeepie #openclaw #worldmonitor

  33. Vibe coding and the open-source Palantir war room 

    ENGLISH] The open-source scene has undergone an incredible transformation in recent years. Since its inception in the 1970s and the rise of Linux, OSS has always been perceived externally as a project centred on freedom, anarchy and the idea of ‘making the world a better place’. It was therefore always met with derision, as the actual revolution – the structural change – went unrecognised. Nor, of course, is this picture entirely accurate: OSS has been used commercially from the very beginning. IBM was one of the movement’s biggest supporters. The entire (commercial) internet is based on OSS; Microsoft is the Linux Foundation’s largest donor, and so on and so forth.

    Since the rise of open-source, open-weight AI models from China, even the tech bros from Stanford and Cambridge, sitting at their Apple laptops, have become absolute devotees of the movement. It is from them that a remarkable statement originates: “Open source has left the realm of philosophy”. This has certainly been true ever since Xi Jinping’s opening speech at the AI Summit in Shanghai (A single string does not make music, and a single tree does not make a forest.).

    But even the classic ‘Power to the People’ OSS and FOSS scene is changing rapidly. Whilst we hear, particularly from Germany and the EU, the pitchfork mobs demanding that OSS projects should not be programmed using AI (‘vibe coding’), Linus Torvalds makes it clear – in essence: ‘Anyone who doesn’t want AI code in the Linux kernel is free to fork it.’ And Keno from Heise Ct says that manual programming will become as exotic as writing letters with a quill pen. It’s ridiculous for the simple reason that anyone using a modern IDE has been using at least autocomplete for years. What do you think that was? A little Chinese man inside the laptop?

    The projects we’re already seeing are fascinating – such as Pewdeepie’s Odyseus or openClaw, or, as shown here in the picture, the OSPalntir war room dashboard WorldMonitor.App, which processes open data from all over the world using AI and makes it accessible to everyone—and can even be run using your own local AI. This is just the beginning.

    word.undead-network.de/2026/08
    #ai #FOSS #heise #keno #linux #oss #Palantir #vibecoding #china #us #usa #future #Pewdeepie #openclaw #worldmonitor

  34. 🚀 Deploy Self-Hosted #OpenClaw on #VPS (3 Minute Quick-Start Guide 🤖)

    This article provides a quick, yet thorough step-by-step guide to deploy self-hosted OpenClaw on VPS servers. A lot of users have been deploying directly to Mac ...
    Continued 👉 #selfhosted #npm #selfhosting #clawdbot

    🚀 Deploy Self-Hosted OpenClaw ...

  35. 🚀 Deploy Self-Hosted #OpenClaw on #VPS (3 Minute Quick-Start Guide 🤖)

    This article provides a quick, yet thorough step-by-step guide to deploy self-hosted OpenClaw on VPS servers. A lot of users have been deploying directly to Mac ...
    Continued 👉 #selfhosted #npm #selfhosting #clawdbot

    🚀 Deploy Self-Hosted OpenClaw ...

  36. Full data: 20 AI agent repos ranked by GitHub stars, OpenRouter daily tokens, npm/PyPI downloads, CVE history, ecosystem size, and Reddit sentiment.

    #Hermes #OpenClaw #Community #SelfHosting #LLM #AI

    glukhov.org/ai-systems/compari

  37. Full data: 20 AI agent repos ranked by GitHub stars, OpenRouter daily tokens, npm/PyPI downloads, CVE history, ecosystem size, and Reddit sentiment.

    glukhov.org/ai-systems/compari

  38. You missed our webinar with #Fastly last week?! 🤯

    No worries, just grab the on-demand version and dive into it at your leisure.

    Learn
    ✓ Why fast performance and strong protection work together
    ✓ How managed #EdgeLayers handle bot traffic that's rarely black and white
    ✓ Using #AI to scale log visibility with #OpenClaw and real-time logging
    ✓ Practical #BotManagement beyond rigid security rules

    Get the recording here
    amazee.io/resources/webinar/se

  39. You missed our webinar with #Fastly last week?! 🤯

    No worries, just grab the on-demand version and dive into it at your leisure.

    Learn
    ✓ Why fast performance and strong protection work together
    ✓ How managed #EdgeLayers handle bot traffic that's rarely black and white
    ✓ Using #AI to scale log visibility with #OpenClaw and real-time logging
    ✓ Practical #BotManagement beyond rigid security rules

    Get the recording here
    amazee.io/resources/webinar/se

  40. OpenClaw или встроенный ИИ?

    Клиент попросил API для подключения OpenClaw и запуска своих ИИ-агентов. Прежде чем делать своё, разобрались вместе: сколько реально стоит поднять агента с нуля (компетенции, доступы, железо) и что модели знают о Directum RX «из коробки».

    habr.com/ru/companies/directum

    #ииагенты #openclaw #генеративные_модели #генеративный_ии #корпоративное_по