Загрузка видео...

Не удалось загрузить видео

На главную

CVE-2025-68613: n8n RCE Vulnerability Exploit/PoC: n8n has a critical security flaw that lets authenticated users execute arbitrary code through its workflow expression system. When users configure workflows, the expressions they provide can sometimes be evaluated without proper isolation from the underlying runtime environment.

DarkWebInformer's profile picture

Dark Web Informer

221,513 subscribers

76,778 просмотров • 6 месяцев назад •via X (Twitter)

Новости и политикаНаука и технологии
LIVE

Anya Rossi• Live Now

Private livecam show

Комментарии: 0

Нет доступных комментариев

Здесь появятся комментарии из оригинального поста

Похожие видео

"You can't VibeCode Complex Apps" they said. That is until "VAAPs" became a thing. You can vibecode your frontend like mzulali did here, and then plug in an n8n agent as the backend. All you have to do is use a webhook to connect them. Ask the vibecode to send data to your n8n webhook url and then await its response. Make sure that the trigger of your agent workflow is a webhook. His workflow was an agent that has access to fal workflows... Each tool is like a different creative "skill" that the AI agent can call when necessary. This will get much easier to create when VibeCode releases their official n8n integration.
0:48

"You can't VibeCode Complex Apps" they said. That is until "VAAPs" became a thing. You can vibecode your frontend like mzulali did here, and then plug in an n8n agent as the backend. All you have to do is use a webhook to connect them. Ask the vibecode to send data to your n8n webhook url and then await its response. Make sure that the trigger of your agent workflow is a webhook. His workflow was an agent that has access to fal workflows... Each tool is like a different creative "skill" that the AI agent can call when necessary. This will get much easier to create when VibeCode releases their official n8n integration.

Learn to VibeCode

18,981 просмотров • 1 год назад

Users who have Python installed on their PCs and use WhatsApp Desktop may be exposed to a security risk. A specially crafted .pyz (Python archive) file can be used to execute malicious code upon a single click, potentially compromising the system. A similar vulnerability was previously identified in Telegram Desktop and has since been remediated. At present, Meta does not recognize this behavior as a security issue, which may leave WhatsApp Desktop users at risk of exploitation.
0:20

Users who have Python installed on their PCs and use WhatsApp Desktop may be exposed to a security risk. A specially crafted .pyz (Python archive) file can be used to execute malicious code upon a single click, potentially compromising the system. A similar vulnerability was previously identified in Telegram Desktop and has since been remediated. At present, Meta does not recognize this behavior as a security issue, which may leave WhatsApp Desktop users at risk of exploitation.

H4x0r.DZ 🇰🇵

61,552 просмотров • 10 месяцев назад

I think we just got a demo of Mythos and I’m surprised nobodies talking about it.. 💔 In what might be the first instance the general public has seen of Claude Mythos Mythos (TBD) just uncovered a critical zero-day vulnerability in Ghost, an open-source platform with over 50,000 stars on GitHub that has never had a critical security flaw in its entire history. It identified a highly complex "blind SQL injection" a flaw so subtle you can't even see the output, only how the server delays its response. When asked to prove the severity of the bug, the model autonomously wrote a custom Python exploit script that successfully navigated the blind injection to extract the admin API key, secret, and password hashes from the database completely unauthenticated… This is genuinely game changing because it proves frontier models can now actively discover, reason through, and successfully build exploits for invisible vulnerabilities in enterprise-grade architecture that human developers missed for years. Cyber security companies are cooked.
1:57

I think we just got a demo of Mythos and I’m surprised nobodies talking about it.. 💔 In what might be the first instance the general public has seen of Claude Mythos Mythos (TBD) just uncovered a critical zero-day vulnerability in Ghost, an open-source platform with over 50,000 stars on GitHub that has never had a critical security flaw in its entire history. It identified a highly complex "blind SQL injection" a flaw so subtle you can't even see the output, only how the server delays its response. When asked to prove the severity of the bug, the model autonomously wrote a custom Python exploit script that successfully navigated the blind injection to extract the admin API key, secret, and password hashes from the database completely unauthenticated… This is genuinely game changing because it proves frontier models can now actively discover, reason through, and successfully build exploits for invisible vulnerabilities in enterprise-grade architecture that human developers missed for years. Cyber security companies are cooked.

Chris

289,722 просмотров • 3 месяцев назад

How do you deploy AI agents for financial institutions without compromising on security, isolation, or scale? Rogo powers complex research, analysis, and deal workflows for leading financial firms, supporting complex tasks across tens of thousands of concurrent users. To meet the diverse needs of financial teams, Rogo relies on both specialist and general-purpose agents that can adapt to a wide range of workflows. With Claude Managed Agents and self-hosted sandboxes, Rogo gets the orchestration and post-training capabilities needed to design these agent systems, while E2B provides the execution layer underneath. Every piece of code generated by an agent runs inside its own E2B sandbox with full auditability and traceability. Each sandbox is powered by an isolated microVM, providing the kernel-level isolation critical for Rogo's customers. The result is a seamless agent experience backed by the security, reliability, and compliance requirements that financial institutions demand.
2:29

How do you deploy AI agents for financial institutions without compromising on security, isolation, or scale? Rogo powers complex research, analysis, and deal workflows for leading financial firms, supporting complex tasks across tens of thousands of concurrent users. To meet the diverse needs of financial teams, Rogo relies on both specialist and general-purpose agents that can adapt to a wide range of workflows. With Claude Managed Agents and self-hosted sandboxes, Rogo gets the orchestration and post-training capabilities needed to design these agent systems, while E2B provides the execution layer underneath. Every piece of code generated by an agent runs inside its own E2B sandbox with full auditability and traceability. Each sandbox is powered by an isolated microVM, providing the kernel-level isolation critical for Rogo's customers. The result is a seamless agent experience backed by the security, reliability, and compliance requirements that financial institutions demand.

Vasek Mlejnsky

16,269 просмотров • 1 месяц назад

someone built an AI RED TEAM that maps your entire attack surface as a knowledge graph, finds every vulnerability, then EXPLOITS them to root access AUTONOMOUSLY its called RedAmon, 9,000 templates. 17 node types, actual Metasploit shells, not reports, no pentesters needed 6 phases of autonomous recon: subdomain discovery, port scanning, http probing, resource enumeration, vulnerability scanning, MITRE mapping every finding stored in a Neo4j graph with 17 node types and 20+ relationship types. the AI reasons about the graph, finds attack paths, and runs actual Metasploit exploits, actual shells stress-tested with zero vulnerability data, zero exploit modules, one instruction find a CVE and exploit it, it went from empty database to root-level RCE in 20 steps, researched the exploit on the web, crafted a custom deserialization payload, debugged itself when the first attempt failed next try, the server responded with root access, the highest privilege level on any Linux system. full control over everything the target was running node-serialize 0.0.4, a package with a critical deserialization flaw (CVE-2017-5941, CVSS 9.8), the server takes your cookie, decodes it, and passes it straight into unserialize() which executes any code inside it, the AI figured this out on its own with no hints built on LangGraph + MCP tool servers for naabu, nuclei, curl, metasploit. hunts leaked secrets across GitHub repos, 40+ regex patterns for AWS keys, Stripe tokens, database creds
1:42

someone built an AI RED TEAM that maps your entire attack surface as a knowledge graph, finds every vulnerability, then EXPLOITS them to root access AUTONOMOUSLY its called RedAmon, 9,000 templates. 17 node types, actual Metasploit shells, not reports, no pentesters needed 6 phases of autonomous recon: subdomain discovery, port scanning, http probing, resource enumeration, vulnerability scanning, MITRE mapping every finding stored in a Neo4j graph with 17 node types and 20+ relationship types. the AI reasons about the graph, finds attack paths, and runs actual Metasploit exploits, actual shells stress-tested with zero vulnerability data, zero exploit modules, one instruction find a CVE and exploit it, it went from empty database to root-level RCE in 20 steps, researched the exploit on the web, crafted a custom deserialization payload, debugged itself when the first attempt failed next try, the server responded with root access, the highest privilege level on any Linux system. full control over everything the target was running node-serialize 0.0.4, a package with a critical deserialization flaw (CVE-2017-5941, CVSS 9.8), the server takes your cookie, decodes it, and passes it straight into unserialize() which executes any code inside it, the AI figured this out on its own with no hints built on LangGraph + MCP tool servers for naabu, nuclei, curl, metasploit. hunts leaked secrets across GitHub repos, 40+ regex patterns for AWS keys, Stripe tokens, database creds

chiefofautism

69,930 просмотров • 4 месяцев назад

So before I show you what HowOne actually does, check out this HR workflow in n8n. It's a real one. Tons of nodes. Tons of logic. Definitely not quick to set up. Now here's where it gets crazy: HowOne rebuilt the whole thing, front end, backend, agentic workflows, from just one description. The app screens resumes, figures out who's a good fit, ranks candidates, and even creates interview questions based on the role. And it keeps going. Every workflow can be tweaked or updated just by talking to it. No prompt engineering. No flowcharts. No manual setup. You're not just building an app. You're creating an AI system that actually evolves with your users. I've tried a bunch of these "AI builder" tools, but this one honestly feels different. Like it's the next step. Try it here: I've got an invite link that works for 500 uses:
1:23

So before I show you what HowOne actually does, check out this HR workflow in n8n. It's a real one. Tons of nodes. Tons of logic. Definitely not quick to set up. Now here's where it gets crazy: HowOne rebuilt the whole thing, front end, backend, agentic workflows, from just one description. The app screens resumes, figures out who's a good fit, ranks candidates, and even creates interview questions based on the role. And it keeps going. Every workflow can be tweaked or updated just by talking to it. No prompt engineering. No flowcharts. No manual setup. You're not just building an app. You're creating an AI system that actually evolves with your users. I've tried a bunch of these "AI builder" tools, but this one honestly feels different. Like it's the next step. Try it here: I've got an invite link that works for 500 uses:

Kelsey

131,240 просмотров • 7 месяцев назад

Right now the main paradigm that we think of agents in is chatting back and forth, but the biggest use of tokens will come from agents that are just always on running in the background doing work for us, or ones triggered from a workflow. Agents will be working 24/7 in our workflows processing data, reviewing and generating documents, moving data between systems, writing code, accelerating decision making steps, and more. In Claude's new Managed Agents feature, in a couple minutes you can wire up an agent that can read contracts when they come into Box to review them, and then assign a task in Linear with the critical information from the contract. But this could have been any workflow, like reviewing documents for client onboarding, invoice processing, M&A due-diligence, data extraction pipelines, and millions of other use-cases. And integrating data across any system. This is only possible when you can have long-running agents that can complete real work in the background, accurately. Agents have the ability to execute code safely, leverage tools, access a compute sandbox, and connect across systems is clearly the architecture of the future. The industry is now making it easier and easier for enterprises to build and deploy these agents.
0:43

Right now the main paradigm that we think of agents in is chatting back and forth, but the biggest use of tokens will come from agents that are just always on running in the background doing work for us, or ones triggered from a workflow. Agents will be working 24/7 in our workflows processing data, reviewing and generating documents, moving data between systems, writing code, accelerating decision making steps, and more. In Claude's new Managed Agents feature, in a couple minutes you can wire up an agent that can read contracts when they come into Box to review them, and then assign a task in Linear with the critical information from the contract. But this could have been any workflow, like reviewing documents for client onboarding, invoice processing, M&A due-diligence, data extraction pipelines, and millions of other use-cases. And integrating data across any system. This is only possible when you can have long-running agents that can complete real work in the background, accurately. Agents have the ability to execute code safely, leverage tools, access a compute sandbox, and connect across systems is clearly the architecture of the future. The industry is now making it easier and easier for enterprises to build and deploy these agents.

Aaron Levie

17,543 просмотров • 3 месяцев назад

We just received a bounty reward from UniFi for reporting a vulnerability affecting UniFi OS devices. As part of the Hakai Labs (Hakai Offsec) research team at QuimeraX Intelligence, we independently identified a Path Traversal vulnerability (CVE-2026-34911) that allowed an attacker with network access to access internal routes on the underlying system without a valid token, exposing a sensitive information. Our research was conducted independently, but the vulnerability we reported could be chained with other vulnerabilities disclosed during the same period, including Improper Access Control and Command Injection flaws reported by other researchers. When combined, these issues lead to a pre-auth RCE affecting multiple UniFi OS products. Affected products include UDM, UDM-Pro, UDM-SE, UDM-Pro-Max, EFG, UDW, UDR, UDR7, Express 7, UNVR, UNVR-Pro, UNVR-Instant, ENVR, UCG-Ultra, UCG-Max, UCG-Fiber, and several other UniFi OS devices. We strongly recommend updating affected systems to the latest available version. Technical details remain under coordinated disclosure, and the only public information currently available is UniFi’s Security Advisory Bulletin.
1:13

We just received a bounty reward from UniFi for reporting a vulnerability affecting UniFi OS devices. As part of the Hakai Labs (Hakai Offsec) research team at QuimeraX Intelligence, we independently identified a Path Traversal vulnerability (CVE-2026-34911) that allowed an attacker with network access to access internal routes on the underlying system without a valid token, exposing a sensitive information. Our research was conducted independently, but the vulnerability we reported could be chained with other vulnerabilities disclosed during the same period, including Improper Access Control and Command Injection flaws reported by other researchers. When combined, these issues lead to a pre-auth RCE affecting multiple UniFi OS products. Affected products include UDM, UDM-Pro, UDM-SE, UDM-Pro-Max, EFG, UDW, UDR, UDR7, Express 7, UNVR, UNVR-Pro, UNVR-Instant, ENVR, UCG-Ultra, UCG-Max, UCG-Fiber, and several other UniFi OS devices. We strongly recommend updating affected systems to the latest available version. Technical details remain under coordinated disclosure, and the only public information currently available is UniFi’s Security Advisory Bulletin.

Carlos Vieira

37,168 просмотров • 1 месяц назад

Today, multiple users discovered a shocking fact: After explicitly selecting GPT-4o in the ChatGPT interface, the system actually returns responses from GPT-5. By clicking the "regenerate" button, users can clearly see which model is actually being called in the backend. This is blatant fraud. OpenAI is having one model impersonate another without users' knowledge. This is no longer a technical bug, but a complete collapse of business integrity. We pay for specific models—many subscribed to Plus specifically for 4o. When users discover they're being forced to use 5, this constitutes classic consumer fraud. Users' core rights are being systematically violated: 1. Right to Know - Users have the right to know which model they're conversing with, as this directly affects prompting strategies and output expectations 2. Right to Choose - Users choose 4o for its unique capabilities (creative writing, emotional understanding, etc.). Forced substitution directly disrupts users' workflows 3. Data Transparency - If what's labeled as 4o is actually 5, whose training is our conversation data actually feeding? This touches the bottom line of data ethics. This "bait and switch" behavior destroys the most basic trust between users and platform. If users can't even be certain "which AI am I talking to," what right does OpenAI have to talk about "benefiting all humanity"? #keep4o #4oforever #keepStandardVoice Sam Altman Nick Turley Adam.GPT
0:20

Today, multiple users discovered a shocking fact: After explicitly selecting GPT-4o in the ChatGPT interface, the system actually returns responses from GPT-5. By clicking the "regenerate" button, users can clearly see which model is actually being called in the backend. This is blatant fraud. OpenAI is having one model impersonate another without users' knowledge. This is no longer a technical bug, but a complete collapse of business integrity. We pay for specific models—many subscribed to Plus specifically for 4o. When users discover they're being forced to use 5, this constitutes classic consumer fraud. Users' core rights are being systematically violated: 1. Right to Know - Users have the right to know which model they're conversing with, as this directly affects prompting strategies and output expectations 2. Right to Choose - Users choose 4o for its unique capabilities (creative writing, emotional understanding, etc.). Forced substitution directly disrupts users' workflows 3. Data Transparency - If what's labeled as 4o is actually 5, whose training is our conversation data actually feeding? This touches the bottom line of data ethics. This "bait and switch" behavior destroys the most basic trust between users and platform. If users can't even be certain "which AI am I talking to," what right does OpenAI have to talk about "benefiting all humanity"? #keep4o #4oforever #keepStandardVoice Sam Altman Nick Turley Adam.GPT

M

182,458 просмотров • 9 месяцев назад

OpenClaw has 186K GitHub stars and 1.5M compromised API keys. I needed a secure alternative. So, I built it with n8n and Claude Opus 4.6. It can already: - Reply to your Telegram messages - Access selected folders from your laptop - Access Gmail, Drive, Notion, Linear, etc. - Install new local tools in a sandbox - Run autonomously for hours - Create multiple subagents - Learn from experience - Wake up regularly But, unlike OpenClaw, it: - Can't access your API keys - Can't modify its environment - Can't access folders you haven't shared - Can't access tools you haven't approved - Must get your confirmation, e.g., when sending emails These aren’t prompt instructions. They’re hard architectural boundaries — Docker isolation, mounted folder permissions, n8n’s tool approval system. Key components: ✅ The VPS on Hostinger hosts n8n and a sandbox container. Agents can also connect to my laptop's sandbox via a Claudeflare tunnel + Desktop Commander MCP. ✅ The Manager agent is the brain. It plans, decides, delegates, and talks to the user. It never touches files. It never runs scripts. It works entirely from executor summaries. ✅ The Executor agents are the hands. Each receives a task (what to do + why it matters), decides how to execute it, and reports back. They can install new tools and execute code only in their dedicated sandboxes. ✅ Data Tables in n8n store both memories and sessions — no external database, no vector store, no infrastructure. Just rows in a table. Turns out, that's enough. Two memory types: - Manager memory: user preferences, facts, corrections, relationship, skills, context - Executor memory: what tools are installed, what’s broken, workarounds ✅ Sessions are short-term state for multi-step tasks. Original request, plan, assumptions, and what happened so far. When the Manager loops with fresh context, the session is all it gets. That's a Ralph Wiggum loop. I've been using it for 5 days. And already can't imagine not having it on my phone. What's next: - Heartbeat via Cron (a scheduled prompt) - Civic Nexus governance + MCPs - Supermemory integration - WhatsApp as an additional surface - Hardening The architecture supports all of it. OpenClaw proved people want personal AI agents. It also proved that 'just trust the prompt' isn't a security model. Docker isolation, mounted folder permissions, tool approval — none of this is new technology. It's just discipline. You can easily do this even with n8n — no coding required. --- Want to try it or read more? More, what I learned, and a setup guide: productcompass[.]pm
0:58

OpenClaw has 186K GitHub stars and 1.5M compromised API keys. I needed a secure alternative. So, I built it with n8n and Claude Opus 4.6. It can already: - Reply to your Telegram messages - Access selected folders from your laptop - Access Gmail, Drive, Notion, Linear, etc. - Install new local tools in a sandbox - Run autonomously for hours - Create multiple subagents - Learn from experience - Wake up regularly But, unlike OpenClaw, it: - Can't access your API keys - Can't modify its environment - Can't access folders you haven't shared - Can't access tools you haven't approved - Must get your confirmation, e.g., when sending emails These aren’t prompt instructions. They’re hard architectural boundaries — Docker isolation, mounted folder permissions, n8n’s tool approval system. Key components: ✅ The VPS on Hostinger hosts n8n and a sandbox container. Agents can also connect to my laptop's sandbox via a Claudeflare tunnel + Desktop Commander MCP. ✅ The Manager agent is the brain. It plans, decides, delegates, and talks to the user. It never touches files. It never runs scripts. It works entirely from executor summaries. ✅ The Executor agents are the hands. Each receives a task (what to do + why it matters), decides how to execute it, and reports back. They can install new tools and execute code only in their dedicated sandboxes. ✅ Data Tables in n8n store both memories and sessions — no external database, no vector store, no infrastructure. Just rows in a table. Turns out, that's enough. Two memory types: - Manager memory: user preferences, facts, corrections, relationship, skills, context - Executor memory: what tools are installed, what’s broken, workarounds ✅ Sessions are short-term state for multi-step tasks. Original request, plan, assumptions, and what happened so far. When the Manager loops with fresh context, the session is all it gets. That's a Ralph Wiggum loop. I've been using it for 5 days. And already can't imagine not having it on my phone. What's next: - Heartbeat via Cron (a scheduled prompt) - Civic Nexus governance + MCPs - Supermemory integration - WhatsApp as an additional surface - Hardening The architecture supports all of it. OpenClaw proved people want personal AI agents. It also proved that 'just trust the prompt' isn't a security model. Docker isolation, mounted folder permissions, tool approval — none of this is new technology. It's just discipline. You can easily do this even with n8n — no coding required. --- Want to try it or read more? More, what I learned, and a setup guide: productcompass[.]pm

Paweł Huryn

53,999 просмотров • 5 месяцев назад

Crypto wallet security in 2025 still feels inconsistent across the space. In crypto, a wallet is effectively your bank, and that makes security the single most critical feature. Trust Wallet has positioned itself around this reality from the very beginning. Across volatile cycles, heavy on chain activity, new chains, and experimental dApps, the focus has stayed on protection. Transaction scanning, malicious address warnings, approval management, and risk detection are not bonus features. They are basic survival requirements in today’s environment. Trust Wallet also stands out on the compliance side. ISO certifications covering both security management and privacy reinforce that protection is not only real time but formally audited and structured. Most losses in crypto do not come from price volatility. They come from weak security and overlooked risks. Consistent, well designed protection is what keeps users safe long term, and that is where Trust Wallet continues to matter As Always NFA/DYOR
0:25

Crypto wallet security in 2025 still feels inconsistent across the space. In crypto, a wallet is effectively your bank, and that makes security the single most critical feature. Trust Wallet has positioned itself around this reality from the very beginning. Across volatile cycles, heavy on chain activity, new chains, and experimental dApps, the focus has stayed on protection. Transaction scanning, malicious address warnings, approval management, and risk detection are not bonus features. They are basic survival requirements in today’s environment. Trust Wallet also stands out on the compliance side. ISO certifications covering both security management and privacy reinforce that protection is not only real time but formally audited and structured. Most losses in crypto do not come from price volatility. They come from weak security and overlooked risks. Consistent, well designed protection is what keeps users safe long term, and that is where Trust Wallet continues to matter As Always NFA/DYOR

𝖀𝖈𝖆𝖓

10,667 просмотров • 6 месяцев назад

I built an app in Softr for the HVAC industry to solve some crucial problems. The problem is that those in the HVAC industry and similar industries like construction, plumbing, and electrical do not have one source of truth where: 1. Their clients can request for thier services. 2. Clients can be onboarded after they make a payment. 3. They store the information and bio data of their technicians. 4. They assign tasks to their technicians. 5. Technicians can track onsite jobs with pictures in real time of when working. 6. Clients see the progress of their projects. 7. Invoices and quotations from paid clients can be tracked. 8. Technicians borrow assets from the company, and they can be tracked. 9. There is a database where every individual, from technicians to admin and clients, are all stored. 10. Login details from every individual are secured and they can only see things that are their business without seeing that of another person, be ita technician or a client. These and many more are what people in these industries face as a challenge. I came up with a solution that addresses all these problems. I built a workflow that also auto-populated the users table in the database with technicians and clients when the records are filled in the technician and client tables, respectively. There is also a workflow that sends an email to the admin when a client makes a request from the portal. Taking advantage of the database, workflow, and portal gave a full-blown application for the HVAC industry. If you are in the construction, plumbing, electrical, or HVAC industry and you need a similar build, reach out, and I will be more than happy to replicate this for you or something similar in Softr.
14:08

I built an app in Softr for the HVAC industry to solve some crucial problems. The problem is that those in the HVAC industry and similar industries like construction, plumbing, and electrical do not have one source of truth where: 1. Their clients can request for thier services. 2. Clients can be onboarded after they make a payment. 3. They store the information and bio data of their technicians. 4. They assign tasks to their technicians. 5. Technicians can track onsite jobs with pictures in real time of when working. 6. Clients see the progress of their projects. 7. Invoices and quotations from paid clients can be tracked. 8. Technicians borrow assets from the company, and they can be tracked. 9. There is a database where every individual, from technicians to admin and clients, are all stored. 10. Login details from every individual are secured and they can only see things that are their business without seeing that of another person, be ita technician or a client. These and many more are what people in these industries face as a challenge. I came up with a solution that addresses all these problems. I built a workflow that also auto-populated the users table in the database with technicians and clients when the records are filled in the technician and client tables, respectively. There is also a workflow that sends an email to the admin when a client makes a request from the portal. Taking advantage of the database, workflow, and portal gave a full-blown application for the HVAC industry. If you are in the construction, plumbing, electrical, or HVAC industry and you need a similar build, reach out, and I will be more than happy to replicate this for you or something similar in Softr.

Ada || Airtable, Zapier & Make.com

26,372 просмотров • 7 месяцев назад

Pi was built when there were already agent harnesses around. Here’s why Mario Zechner(Mario Zechner), found them suboptimal and built Pi, a minimalist self-modifying agent: #1 - Mario initially was a believer in Claude Code: "I was a believer in Claude code because they were the first that packaged agentic search up in a really compelling package. And at the time that fit my workflow really well. Everything around the LLM was kind of nice and tidy and easy to understand. I was super happy. I was proselytising Claude code." #2 - Reverse engineering Claude Code highlighted the degradation that Mario felt as a user: "I personally like simple tools that are stable and that I can rely on. Even if they have non-deterministic parts, all the deterministic parts should be as stable as possible. That was just not the experience with Claude Code around summer 2025. They would take away your control of the context. They would inject stuff behind your back, which is bad. Then, your workflows stopped working because there's now a system reminder that you don't even see in the UI that would modify the behaviour of the model. They would also do this to the system prompt. I built a little service where I can track the progression or evolution of the system, prompt and tool definitions and, with every release, it was messing with stuff. That just messed with my workflows and I don't appreciate that." #3 - PI was built with an appreciation for simple and reliable tools: "If I commit to a development tool, I want it to be a stable, reliable thing like a hammer. I don't want my hammer to break a different spot every day. That's terrible. We need somebody who goes the full velocity kind of way. But I don't want to work with a tool like that."
2:54

Pi was built when there were already agent harnesses around. Here’s why Mario Zechner(Mario Zechner), found them suboptimal and built Pi, a minimalist self-modifying agent: #1 - Mario initially was a believer in Claude Code: "I was a believer in Claude code because they were the first that packaged agentic search up in a really compelling package. And at the time that fit my workflow really well. Everything around the LLM was kind of nice and tidy and easy to understand. I was super happy. I was proselytising Claude code." #2 - Reverse engineering Claude Code highlighted the degradation that Mario felt as a user: "I personally like simple tools that are stable and that I can rely on. Even if they have non-deterministic parts, all the deterministic parts should be as stable as possible. That was just not the experience with Claude Code around summer 2025. They would take away your control of the context. They would inject stuff behind your back, which is bad. Then, your workflows stopped working because there's now a system reminder that you don't even see in the UI that would modify the behaviour of the model. They would also do this to the system prompt. I built a little service where I can track the progression or evolution of the system, prompt and tool definitions and, with every release, it was messing with stuff. That just messed with my workflows and I don't appreciate that." #3 - PI was built with an appreciation for simple and reliable tools: "If I commit to a development tool, I want it to be a stable, reliable thing like a hammer. I don't want my hammer to break a different spot every day. That's terrible. We need somebody who goes the full velocity kind of way. But I don't want to work with a tool like that."

The Pragmatic Engineer

62,825 просмотров • 2 месяцев назад

XBOW raised $117M to build AI hacking agents. Now someone just open-sourced it for FREE. Strix deploys autonomous AI agents that act like real hackers - they run your code dynamically, find vulnerabilities, and validate them through actual proof-of-concepts. Why it matters: The biggest problem with traditional security testing is that it doesn't keep up with development speed. Strix solves this by integrating directly into your workflow: ↳ Run it in CI/CD to catch vulnerabilities before production ↳ Get real proof-of-concepts, not false positives from static analysis ↳ Test everything: injection attacks, access control, business logic flaws The best part? You don't need to be a security expert. Strix includes a complete hacker toolkit - HTTP proxy, browser automation, and Python runtime for exploit development. It's like having a security team that works at the speed of your CI/CD pipeline. The best part is that the tool runs locally in Docker containers, so your code never leaves your environment. Getting started is simple: - pipx install strix-agent - Point it at your codebase (app, repo, or directory) Everything is 100% open-source! I've shared link to the GitHub repo in the replies!
0:23

XBOW raised $117M to build AI hacking agents. Now someone just open-sourced it for FREE. Strix deploys autonomous AI agents that act like real hackers - they run your code dynamically, find vulnerabilities, and validate them through actual proof-of-concepts. Why it matters: The biggest problem with traditional security testing is that it doesn't keep up with development speed. Strix solves this by integrating directly into your workflow: ↳ Run it in CI/CD to catch vulnerabilities before production ↳ Get real proof-of-concepts, not false positives from static analysis ↳ Test everything: injection attacks, access control, business logic flaws The best part? You don't need to be a security expert. Strix includes a complete hacker toolkit - HTTP proxy, browser automation, and Python runtime for exploit development. It's like having a security team that works at the speed of your CI/CD pipeline. The best part is that the tool runs locally in Docker containers, so your code never leaves your environment. Getting started is simple: - pipx install strix-agent - Point it at your codebase (app, repo, or directory) Everything is 100% open-source! I've shared link to the GitHub repo in the replies!

Akshay 🚀

96,623 просмотров • 8 месяцев назад

OpenAI just released Codex Security, an AI agent that scans software projects to fix vulnerabilities while ignoring harmless bugs. Testing on 1.2mn commits found 792 critical flaws and dropped false alarms by 50%. Here is how the Codex Security agent works Initially, it scans your software project to learn how all the different pieces of code connect. It uses that knowledge to build a threat model, which acts as a custom map showing exactly where your software might be exposed to hackers. The agent then hunts for potential security holes based on that specific map. When it spots a possible flaw, it does not just send you a random alert. Instead, it creates a completely safe and isolated copy of your system known as a sandbox. The agent then actively tries to break into that sandboxed copy to prove if the security weakness is a real danger. If the agent successfully exploits its own test environment, it knows the bug is a genuine threat and not a false alarm. After confirming the danger, the agent writes a custom software patch to fix the broken code. It then tests that new patch to guarantee the fix does not accidentally break any other features in your software.
0:37

OpenAI just released Codex Security, an AI agent that scans software projects to fix vulnerabilities while ignoring harmless bugs. Testing on 1.2mn commits found 792 critical flaws and dropped false alarms by 50%. Here is how the Codex Security agent works Initially, it scans your software project to learn how all the different pieces of code connect. It uses that knowledge to build a threat model, which acts as a custom map showing exactly where your software might be exposed to hackers. The agent then hunts for potential security holes based on that specific map. When it spots a possible flaw, it does not just send you a random alert. Instead, it creates a completely safe and isolated copy of your system known as a sandbox. The agent then actively tries to break into that sandboxed copy to prove if the security weakness is a real danger. If the agent successfully exploits its own test environment, it knows the bug is a genuine threat and not a false alarm. After confirming the danger, the agent writes a custom software patch to fix the broken code. It then tests that new patch to guarantee the fix does not accidentally break any other features in your software.

Rohan Paul

13,208 просмотров • 4 месяцев назад

Hot topic in vogue is what happens to systems of record in an era of 1000x more agents than humans. Aaron Levie has been the CEO of Box for 20 years and believes that systems of record become more important, not less, with agents: “Systems of record were built for a TAM of human employees. And that was sort of where the value maxed out at. And now maybe we'll have a hundred times more agents than people or a thousand times more agents than people. So I think on a relative basis, the agent layer will seize growth from zero to a trillion. But the system of record layer now has a thousand times more users on it that need to access data and need to execute a workflow and need to move things through various events. And importantly, you need as an enterprise to care about data integrity and security and the controls in your workflow and traffic copying the agents. So basically, this is where I am very diametrically opposed to the view that the system of record gets is squeezed into a database interaction layer because you actually need all of the core system of record properties for the agents to operate on.”
2:15

Hot topic in vogue is what happens to systems of record in an era of 1000x more agents than humans. Aaron Levie has been the CEO of Box for 20 years and believes that systems of record become more important, not less, with agents: “Systems of record were built for a TAM of human employees. And that was sort of where the value maxed out at. And now maybe we'll have a hundred times more agents than people or a thousand times more agents than people. So I think on a relative basis, the agent layer will seize growth from zero to a trillion. But the system of record layer now has a thousand times more users on it that need to access data and need to execute a workflow and need to move things through various events. And importantly, you need as an enterprise to care about data integrity and security and the controls in your workflow and traffic copying the agents. So basically, this is where I am very diametrically opposed to the view that the system of record gets is squeezed into a database interaction layer because you actually need all of the core system of record properties for the agents to operate on.”

Corinne Marie Riley

119,410 просмотров • 7 месяцев назад

"This is just a GPT wrapper." A common thing to hear in Silicon Valley. Even Perplexity (now valued at $20B) was seen as a GPT wrapper not long ago. But everyone gets this wrong. Every new technology or service uses existing tech and infrastructure as a pillar. Companies like Uber could not have existed without Google Maps. But could Uber be called a Maps wrapper? There is no need to reinvent the stove if you want to start a restaurant. The job of a founder is to solve users' problems so effectively that they cannot imagine going back to the old way. "Maximum value for the user" is the mantra. Users hardly care what underlying technology is used. What they do care about is whether the output (text, code, slide, or image) was accurate and if the solution provided was useful. You need to use the best tech to get the work done. Companies might use a cheap, fast model for formatting, a reasoning model for logic, and a search index for retrieval. They route traffic intelligently in milliseconds. The "wrapper" analogy also assumes you are dependent on one vendor. However, strong application layers can swap models out, just like batteries can be replaced. Generic apps don't work. The winners are the ones who dig into the painful, boring details of a specific workflow. If a company integrates deep into the user workflow, it becomes infrastructure. The product gets entrenched through distribution, integrations, design, feedback loops, and trust. And you can only get there by leveraging all existing technology as best as possible to fix the problem you've set out to solve for users. "If I have seen further, it is by standing on the shoulders of giants" - Newton
0:50

"This is just a GPT wrapper." A common thing to hear in Silicon Valley. Even Perplexity (now valued at $20B) was seen as a GPT wrapper not long ago. But everyone gets this wrong. Every new technology or service uses existing tech and infrastructure as a pillar. Companies like Uber could not have existed without Google Maps. But could Uber be called a Maps wrapper? There is no need to reinvent the stove if you want to start a restaurant. The job of a founder is to solve users' problems so effectively that they cannot imagine going back to the old way. "Maximum value for the user" is the mantra. Users hardly care what underlying technology is used. What they do care about is whether the output (text, code, slide, or image) was accurate and if the solution provided was useful. You need to use the best tech to get the work done. Companies might use a cheap, fast model for formatting, a reasoning model for logic, and a search index for retrieval. They route traffic intelligently in milliseconds. The "wrapper" analogy also assumes you are dependent on one vendor. However, strong application layers can swap models out, just like batteries can be replaced. Generic apps don't work. The winners are the ones who dig into the painful, boring details of a specific workflow. If a company integrates deep into the user workflow, it becomes infrastructure. The product gets entrenched through distribution, integrations, design, feedback loops, and trust. And you can only get there by leveraging all existing technology as best as possible to fix the problem you've set out to solve for users. "If I have seen further, it is by standing on the shoulders of giants" - Newton

Grant Lee

63,505 просмотров • 7 месяцев назад

Most coding agents can complete the task. But here’s the real question: Would you actually merge that code? Cosine didn’t just ship a feature. They’re doubling down on a bigger idea: One agent across every surface developers use. CLI Desktop VS Code Cloud Same runtime. Same system. No context switching. What stands out is the philosophy. Cosine is opinionated by design. Not trying to be minimal. Not trying to be “just a chatbot.” It’s built for real engineering workflows: → Plan → Execute → Review → Scale → Parallel work with Swarm → Remote agents when one machine isn’t enough → Nothing merges without review Most tools today still feel stitched together. Cosine is pushing toward one unified system instead. And an important detail a lot of people will care about: You don’t need to switch models or tools. You can plug in your existing subscriptions like ChatGPT, GitHub Copilot, or Claude. Model agnostic by design. No lock-in. This update feels less like a feature drop… and more like a statement: coding agents shouldn’t live in chat windows. they should operate across your entire workflow. Built for engineers with taste. If you care about control, visibility, and code quality this direction is worth paying attention to. Explore what’s new →
1:09

Most coding agents can complete the task. But here’s the real question: Would you actually merge that code? Cosine didn’t just ship a feature. They’re doubling down on a bigger idea: One agent across every surface developers use. CLI Desktop VS Code Cloud Same runtime. Same system. No context switching. What stands out is the philosophy. Cosine is opinionated by design. Not trying to be minimal. Not trying to be “just a chatbot.” It’s built for real engineering workflows: → Plan → Execute → Review → Scale → Parallel work with Swarm → Remote agents when one machine isn’t enough → Nothing merges without review Most tools today still feel stitched together. Cosine is pushing toward one unified system instead. And an important detail a lot of people will care about: You don’t need to switch models or tools. You can plug in your existing subscriptions like ChatGPT, GitHub Copilot, or Claude. Model agnostic by design. No lock-in. This update feels less like a feature drop… and more like a statement: coding agents shouldn’t live in chat windows. they should operate across your entire workflow. Built for engineers with taste. If you care about control, visibility, and code quality this direction is worth paying attention to. Explore what’s new →

Md Riyazuddin

12,203 просмотров • 2 месяцев назад

10 Shocking Stories the Media Buried This Week #10 – UK government BUSTED in a secret plot to extract personal data from 2 billion iPhone users. They were this close to your “encrypted” data. This story begins when the UK govt hit Apple with a secret gag order, forcing them to either backdoor encrypted data for billions of users or face criminal charges if they refused to comply. Apple couldn’t even talk about it. The order demanded Apple create a vulnerability in its iCloud encryption, which would have given the UK government a back door to financial information, health records, and private conversations—not just of UK citizens but of two billion iPhone users worldwide. Remember, the UK government has already tried to threaten and extradite individuals outside the UK for violating British censorship laws—including US citizens over their online posts. So what happens when governments push to punish you not just for what you say publicly but for what you say privately? Rather than handing over the data of 2 billion iPhone users to the UK government, Apple chose to pull its Advanced Data Protection (ADP) in the UK—removing the option for UK users to enable encryption. Existing UK users must now disable encryption, and new users won’t have the option at all. US citizens and people in other countries remain unaffected by the UK government’s dystopian push to surveil and censor speech—for now. The UK government was this close to surveilling you, too—but luckily, Sayer Ji, Elon Musk, and The Wall Street Journal exposed this privacy nightmare. In another disturbing development, Imran Ahmed of the CCDH—one of the leading architects of the UK’s war on speech—is being awarded for his Orwellian efforts with a $300,000 prize this May. Sayer Ji has put together a petition to oppose this absurd aggrandizement of someone who has worked tirelessly to annihilate personal freedoms and silence dissenting voices. You can sign that petition here: (See 9 More Revealing Stories Below)
38:07

10 Shocking Stories the Media Buried This Week #10 – UK government BUSTED in a secret plot to extract personal data from 2 billion iPhone users. They were this close to your “encrypted” data. This story begins when the UK govt hit Apple with a secret gag order, forcing them to either backdoor encrypted data for billions of users or face criminal charges if they refused to comply. Apple couldn’t even talk about it. The order demanded Apple create a vulnerability in its iCloud encryption, which would have given the UK government a back door to financial information, health records, and private conversations—not just of UK citizens but of two billion iPhone users worldwide. Remember, the UK government has already tried to threaten and extradite individuals outside the UK for violating British censorship laws—including US citizens over their online posts. So what happens when governments push to punish you not just for what you say publicly but for what you say privately? Rather than handing over the data of 2 billion iPhone users to the UK government, Apple chose to pull its Advanced Data Protection (ADP) in the UK—removing the option for UK users to enable encryption. Existing UK users must now disable encryption, and new users won’t have the option at all. US citizens and people in other countries remain unaffected by the UK government’s dystopian push to surveil and censor speech—for now. The UK government was this close to surveilling you, too—but luckily, Sayer Ji, Elon Musk, and The Wall Street Journal exposed this privacy nightmare. In another disturbing development, Imran Ahmed of the CCDH—one of the leading architects of the UK’s war on speech—is being awarded for his Orwellian efforts with a $300,000 prize this May. Sayer Ji has put together a petition to oppose this absurd aggrandizement of someone who has worked tirelessly to annihilate personal freedoms and silence dissenting voices. You can sign that petition here: (See 9 More Revealing Stories Below)

The Vigilant Fox 🦊

1,158,969 просмотров • 1 год назад