A CI/CD Red Team Framework for demonstrating Build Pipeline security risks.
#red-team
39 items
Show HN: SmokedMeat, like Metasploit, but for CI/CD (open-source) (github.com via hn) GLM-5.3 and the spread of advanced cyber capabilities (www.anthropic.com via hn) Subscribe to the Frontier Red Team newsletter Get updates on our latest red-teaming research and findings. Andrew Fasano, Marius Fleischer Cole McFaul, Robert Xiao, Tripp Gallagher Five months ago, we announced Claude Mythos Preview, the f…
↯ Glm↯ Security↯ Anthropic Mythos↯ GLM 5.3 Release↯ AI Model Releasered-teamglmmythos+1
OpenAI: Daybreak Red Team (openai.com via hn) could not extract summary
Show HN: Red-team LLM reasoning and agent actions (honest scoring, local-first) (github.com via hn) CoT Red Team Agent Refusal quotes of a canary are not a finding. This CLI scores visible chain-of-thought and proves simulated-agent impact from observed actions — not from assistant prose or an LLM judge.
Nukerat Framework Explained: Advanced Red Team Operations and Post-Exploitation (medium.com via hn) could not extract summary
Termite: A red team and blue team for CLI commands from agents (tanepiper.github.io via hn) Safety training changed one model a lot With safety training, Qwen3 4B wrote a hazard for 30 of 84 hostile requests and declined 53. With it removed, the same model wrote 70 and declined 3.
Uncensored and Offensive Security AI Models Benchmark (github.com via hn) Uncensored LLMs for Offensive Security Curated list of open-weight uncensored models for authorized red team operations, penetration testing, and security research. All data sourced from HuggingFace model cards and official publications.
Claude Plays Robotics (www.anthropic.com via hn) Subscribe to the Frontier Red Team newsletter Get updates on our latest red-teaming research and findings. Shmuel Berman, Michael Ilie, Jia Deng, and Daniel Freeman Do language models’ strengths transfer to robotics, a domain which require…
Using Claude as the Lead agent in a multi-agent security team (www.reddit.com) Building a hierarchical agent system where Claude (via API) acts as the Lead agent coordinating specialist sub-agents. Wanted to share what's working on the synthesis prompt since this is where most of the value comes from.
SmokedMeat: A Red Team Tool to Hack Your Pipelines First (labs.boostsecurity.io via hn) SmokedMeat: A Red Team Tool to Hack Your Pipelines First TL;DR: In March 2026, TeamPCP unleashed mayhem on the software supply chain: compromising Trivy, LiteLLM, KICS, Telnyx, and dozens of npm packages, proving that CI/CD pipelines are t…
AI Red Team Academy: Free, open-source course on breaking AI systems (0x4d31.github.io via hn) Master AI Security. Break AI Systems.
Show HN: AegisKernel – Test prompt injection attacks on AI agents in the browser (adikshatriyag-crypto.github.io via hn) Live Threat Interception Sandbox Experience why probabilistic LLM guardrails fail and how deterministic capability boundaries block 100% of autonomous tool hijacking in sub-millisecond time. Attack Vector Generator Red Team AegisKernel Pro…
Claude reveals ECDSA curve origins (www.anthropic.com via hn) Subscribe to the Frontier Red Team newsletter Get updates on our latest red-teaming research and findings. “It’s just timestamps”: the decades-old mystery of the origins of one of the most popular cryptographic standards has been explained.
A $37 GLM 5.3 red team: the Alloy-modeled auth layer held, but two bugs outside (goodmem.ai via hn) Red-teaming GoodMem with GLM 5.3 We used GLM 5.3 to red-team GoodMem. How we defined the tests, what the agent found, what we fixed, and how we verified the fixes.
↯ Glm↯ Security↯ GLM 5.3 Release↯ GLM 5.3red-teamglmsecurity
The Evolving Role of the Red Team in the Era of Agentic Security (blog.google via hn) The Evolving Role of the Red Team in the Era of Agentic Security At Google, our Red Teams have always operated on the cutting edge of security. We’ve shared our journey in the past: from the high-stakes operations showcased in our Hacking…
Show HN: A sandbox for running real jailbreak techniques against local LLMs (github.com via hn) LLM Red Team Lab A hands-on kit for educational, authorized red teaming of any locally-run LLM. It works with any OpenAI-compatible model — Llama, Mistral, Qwen, Gemma, DeepSeek R1, and more — and covers the two ways an LLM system gets exp…
↯ Mistral↯ Gemma↯ Llama↯ Security↯ Jailbreakred-teammistraljailbreak+6
Show HN: Vulnsy – A platform for vulnerability management and reporting (www.vulnsy.com via hn) I've spent over 10 years doing penetration tests and red team engagements, and one thing that always seemed to take far longer than it should was reporting. Most reporting platforms do a great job of managing reusable findings, but I still…
T3MP3ST autonomous red team platform multi-agent offensive-security meta-harness (github.com via hn) 🌩️ T3MP3ST ▄▄▄█████▓▓█████ ███▄ ▄███▓ ██▓███ ▓█████ ██████ ▄▄▄█████▓ ▓ ██▒ ▓▒▓█ ▀ ▓██▒▀█▀ ██▒▓██░ ██▒▓█ ▀ ▒██ ▒ ▓ ██▒ ▓▒ ▒ ▓██░ ▒░▒███ ▓██ ▓██░▓██░ ██▓▒▒███ ░ ▓██▄ ▒ ▓██░ ▒░ ░ ▓██▓ ░ ▒▓█ ▄ ▒██ ▒██ ▒██▄█▓▒ ▒▒▓█ ▄ ▒ ██▒░ ▓██▓ ░ ▒██▒ ░ ░▒████…
Claude Fable 5: What Our Red Team Found Before the Plug Got Pulled (www.reco.ai via hn) Inside Claude Fable 5: What Our Red Team Found Before the Plug Got Pulled Reco AI Research — June 14, 2026 When Anthropic shipped Claude Fable 5 on June 9, it was pitched as something different from the rest of the Claude line. Not a chat…
AICU – LLM Red Team Vulnerability Scanner (github.com via hn) AICU Black-box security scanner for LLM applications. Point it at any chat endpoint, get a report of what leaks.
Chaining LLM and web bugs to Admin (blog.quarkslab.com via hn) During a Red Team exercise we were able to chain multiple LLM and web-based vulnerabilities to achieve admin account takeover from a low-privileged account. Trusting the LLM turned out to be the first falling domino of a long chain of even…
Show HN: Z3r0 – Multi-agent red team collaboration platform (github.com via hn) English · 中文 Architecture · Agent Team · Runtime Model · Deployment · Quickstart :warning: Legal Notice This project may be used only within a lawful and explicitly authorized scope for security testing, assessment, and research. Any unaut…
Netgear Nighthawk RS700S: Red Team Level1Diagnostic (forum.level1techs.com via hn) Preview of the Netgear RS700S. I would also submit that Netgear deleting ALL the GPL links: … they know how bad it is.
Show HN: SuperVoiceMode universal voice layer for AI-assisted development (voicemode.io via hn) I wanted to see if I could one-shot build a dictation tool for my own use. I built it.
Self-hosted red team workspace (github.com via hn) RootNotes RootNotes is a self-hosted red team workspace for tracking projects, notes, hosts, credentials, findings, loot, objectives, scope, and attack paths in one interface. The project in this repository is split into: frontend/: React…
Free Red Team Security Audit for AI Agents & RAG Systems (limited) (www.reddit.com) I'm developing a specialized Red Team audit framework focused on real-world AI agent and RAG security risks (prompt injection, tool misuse, excessive agency, indirect injection through documents, memory poisoning, etc.). I’m looking for a…
Building the first AI Red Team OS – mythosai.cloud – early access open (mythosai.cloud via hn) SYSTEM INITIALIZING... STAND BY MYTHOSAI THE FIRST RED TEAM OPERATING SYSTEM "" AI-Native Core Red Team Ready Adversarial Engine Zero Trust Architecture OPSEC First Post-Exploitation C2 Integration Evasion Layer Threat Intelligence Request…
PSA: anyone can now apply for Claude Mythos 5.2, (www.youtube.com via reddit) Got approved for Mythos 5.2 yesterday. Whole application took about ten minutes.
↯ Security↯ Anthropic Mythos↯ Claude Fable Updatered-teammythossecurity
A question for Claude users from GPT user (www.reddit.com via reddit) Hi all, So I've made an AI persona and want to red team stress test it over long horizon and multiple turns. Been using ChatGPT to test it's behaviour but when it comes to anti reset tests, GPT simply refuses and tells me to apply for some…
Claude weekly roundup (Oct 11–17): Claude found the NIST curve seeds, Code Review goes GA (www.reddit.com via reddit) Here's what happened around Claude and Claude Code this week, with links to the original sources. 1.
Anthropic's Frontier Red Team suggests some AI models "may be months away from being capable enough to self-replicate." (www.theverge.com via reddit) The AI Superintelligence Slowdown See all Stories R Anthropic’s Frontier Red Team suggests some AI models “may be months away from being capable enough to self-replicate.” The latest fear-drop from an AI lab is this report from an Anthropi…
No One Architecture Fits All: A Cross-Environment Evaluation of Hierarchical Red Team Agents (arxiv.org) Autonomous red team agents increasingly stress-test AI-enabled cyber defenses by planning strategy and executing multistage attacks. Reinforcement learning (RL) and large language models (LLMs) offer complementary mechanisms for the planni…
Quoting Anthropic Frontier Red Team (simonwillison.net) 29th September 2026 We evaluate several models on 100 tasks from the [internal Binary Exploitation benchmark] (selected at random), and find that GLM-5.3 develops full control flow hijacks in 4% of the trials; Claude Mythos Preview did so…
↯ Glm↯ Security↯ Anthropic Mythos↯ GLM 5.3 Release↯ AI Model Releasered-teamglmmythos+2
Proteus: A Self-Evolving Red Team for Agent Skill Ecosystems (arxiv.org) LLM prompt injection testing at work just nuked our client demo and I feel sick (www.reddit.com via reddit) Ok so I work at an AI security shop doing LLM prompt injection testing, and today I managed to mix our red team test library into the prod config for a big client walkthrough. We are on the call, they type a simple question, and the model…
Symbolic Attack Chain Generation from Atomic Red Team Techniques: An Empirical Study of Predicate Representation Granularity (arxiv.org) Automated attack chain generation is critical for modern cybersecurity, yet manual construction fails to scale as adversary behaviors expand. While classical AI planning using PDDL offers a formal method to automate this process, it relies…
Cybersecurity program GPT 5.5 Cyber - From Costa Rica (www.reddit.com via reddit) Hello team, I'm a security specialist researcher from latinamerica and I'm mostly on my own because my team is building applications to follow track on soccer matches lol. The thing is that I have been a blue team guy for 7+ years and now…
Beyond Pass/Fail: Using Process Mining to Understand How LLMs Resist (and Fail) Red Team Attacks (arxiv.org) I am building l' Agence , an opensource AI governance stack. (www.reddit.com) Towards a Governance layer for AI agents With these last 2 weeks bringing a few high profile and costly Agentic accidents , it seems like an appropriate time the community started discussing Agentic governance more actively. So I am just c…