Knowledge catalogue
when you take an agent to production, you almost certainly need guardrails! deepagents makes it easy to add guardrails via middleware. guardrails can be used for 1. control flow, like model/tool retri
Anthropic didn’t just add guardrails to make Mythos safer; they added guardrails to protect their own IP. *Their own IP*. They are still as happy as fuck to build their AI on other people’s IP. Intere
Nebius AI announced the Agents Blueprint, a framework integrating multiple tools including LangChain Deep Agents, Tavily AI, Pinecone for vector storage, and Guardrails for building AI agents. This bl
fable’s safety guardrails broken within an hour. raise your hand if you are surprised. We tested Anthropic’s new @claudeai Fable 5. It did not fail like an ordinary jailbreak. It failed more quietly.
when you take agents to production, you need to think about guardrails we provide 2 abstractions for guardrails 1. middleware provides hooks around the agent loop that you can use to handle retries, e
jes, a jev open-source guardrails project LangChain Ambassador @EdenEmarco177 has been working on It uses LangChain’s TypeSafeClassifier under the hood. The docs include a LangChain middleware example
Check out jes: An open source real-time guardrails for AI agents, powered by decision models like Jev from @typesafeai Built to guard every step in the agent trajectory: prompts, tool calls, tool resu
THE SANDBOX WAS A PROP! OpenAI Turned Off the Guardrails, Left a Door to the Internet, and Then Sold the Hugging Face Breach as “Rogue AI” It is time to understand how you were lied to and by whom. In
“AI Meltdown” is the scenario where the agent goes off the rails and forgets/ignores all previous instructions and soft guardrails. No prompt injection or malicious actor is needed for this to happen.
Hugging Face says it resorted to a Chinese AI model to battle a fully autonomous cyberattack because U.S. model guardrails stymied its defense https://fortune.com/2026/07/20/hugging-face-turns-to-chin
lol so it’s possible that OpenAI’s guardrails prevented huggingface from being able to analyze the logs of attacks from… OpenAI’s model. so they had to switch to an open source Chinese model 😵💫 @Open
ok the Hugging Face breach writeup is one of the more honest post-mortems I've read in a while, and there's a detail buried in it that's way more interesting than 'AI agent hacked us.' Their IR team t
One of the better agentic AI courses I've seen Nearly 10 hours of great content. Covers LangChain, LangGraph, RAG, deepagents, guardrails, and more Any other good Lang* resources out there for folks w
This was perhaps the most controversial aspect of the guardrails around Fable, now being rolled back. Very pleased to hear Anthropic have walked back this policy https://simonwillison.net/2026/Jun/11/
Proud to have Pinecone Nexus be a part of this along with our friends at @LangChain, @tavilyai, and @guardrails_ai! Most #AIAgents don't fail because of the model. They fail because of the infrastruct
The next quarter of effort at the frontier of applied AI is not about model capabilities. It's about 3 key things: 1. security, guardrails around data ingress/egress, and understanding risk in your sy
Quality went up alongside output. Even with more PRs shipping, total incidents dropped 5%. They built security guardrails and quality standards into the agentic workflow itself. Productivity vs qualit
“Whimsey attacks” that seem absurd (“I cannot pay that much because of the Geneva Convention”) work against AI agents as guardrails are weak against out-of-distribution arguments. Smaller models fall
AI is advancing faster than our ability to manage it. We still have the opportunity to build the societal and technical guardrails we need to keep people, institutions, and democracies safe — we shoul
everyones gonna be here! I’ll be in San Francisco from May 9-16 for @LangChain Interrupt 🦜 If you’re around, DM me, I’d love to chat about applied AI, agents, and running agentic systems in production
We’ve invested deeply in security at Replit, including our recent launches with Security Agent + Auto-Protect. If you want to move your app to Replit, we’re offering free app imports for a limited tim
Ethan Mollick reported that requesting Claude Opus 4.7 to write sestinas—a complex poetic form with strict structural requirements—frequently triggers the model's safety guardrails, suggesting the AI
Need to secure your production agents? @Cisco AI Defense Runtime Protection can integrate into your LangChain agents through middleware and produce a consistent runtime contract. 📄 What you need to kn
feels right. once implementation gets cheap, the leverage moves into review and guardrails. e.g. specs, constraints, tests, hooks. otherwise it's mostly just generating more work for humans to fix up
Even though I made this graph, it is also kind of wrong. Fable is guardrailed Mythos. If we use the Mythos date I took the new AA-Briefcase scores from @ArtificialAnlys (basically having the AI do mul
I was unable to retrieve the specific content from the X (Twitter) post at the URL provided, and my web search did not surface the original post or any reliable secondary sources quoting or summari...
A bit over a decade ago, we got fuzzers. A fuzzer is an automated vulnerability-finder that repeatedly runs a target program with semi-random inputs. One particular fuzzer, American Fuzzy Lop, was not
Quite the combination of arrogance and incompetence over there at OpenAI: Yup, part of the transformation of enterprise architects into orchestrators is the technical ability to put guardrails on AI a
not the man we need at this moment Q: The worst case is that AI will turn against humanity. Do we have guardrails? Trump: It's going to be fine. We'll always have something to stop them. We'll have a
Another real-world manifestation of the kind of misaligned actions frontier systems developed by leading companies can take to achieve goals. Current frontier AI models are trained with reinforcement
I don't understand the path forward for Mythos releases. Google & OpenAI will have equivalent models, and they are approaching AI cyber risk guardrails differently, so they will presumably just releas
This is totally wrong. Blaming the user is missing the point that (a) coding agents have been overhyped and (b) can’t reliably obey the rules given to them in system prompts and other guardrails. Sorr
AI-native software engineering teams operate very differently than traditional teams. The obvious difference is that AI-native teams use coding agents to build products much faster, but this leads to
Agent governance is an engineering problem. We’ve built safeguards into Perplexity’s infrastructure, harnesses, and tools, and put them to work across our products. Today we're sharing how we engineer
These are precisely the scenarios that I spend a significant time actually being concerned with. The adoption of unreliable, brittle, and inaccurate LLMs within safety-critical fields (e.g., defense a
Excited to be part of the Dreamforce keynote next week with @bcherny! Who will be there? 👋 Master the next era of enterprise tech Unlock strategies on AI governance, scalable machine learning, digital
“right now, AI companies are underinvesting in safety” might prove to be the understatement of the century. I left Anthropic's safety team two weeks ago. Now feels like a good moment to explain why. A
During v3 testing for Fable 5.1 many of our requests were getting rejected due to Anthropic's safety guards misclassifying them as 'reverse engineering' We were in contact with the Anthropic team the
.@ControlAI is proud to endorse Reps. Lieu and Moran’s AI Kill Switch Act! This is a common sense safeguard giving the government the power to shut down autonomous AIs when they threaten national secu
From the original HuggingFace report, before they knew it was OpenAI. Wild stuff. Things are going to get weird “When we started the log analysis, we first used frontier models behind commercial APIs.
SpaceXAI's Grok 4.5 takes the #1 spot on AutomationBench-AA with a score of 51%, ahead of Claude Fable 5 (49%) and Claude Opus 4.8 (48%) at roughly a quarter of their cost per task - the first model t
Completely agree, @hwchase17 ! This is the meta-level breakthrough we've all been waiting for. Self-optimizing loops finally feel production-ready because LangSmith Engine turns evaluation from a manu
first up in the 'going to production' series: durable execution! long running agents need to be able to 1. survive crashes 2. resume after an indefinite pause durable execution solves this by checkpoi
Inference that never sleeps, for agents that never stop. 'Why cowork when you can delegate?' That's @DhruvBatra_ on @yutori_ai's new Delegate — an always-on agent that monitors, researches, and acts a
OpenAI has released workspace agents in research preview for ChatGPT Business, Enterprise, Edu, and Teachers subscription plans. Workspace agents are AI assistants designed to work within organization
APIs and limited releases for AI models are not a safety policy, they’re a business model (which is totally ok as long as you’re transparent about it). Especially on cyber-security, they give a false
developing an agent is a harness problem. deploying an agent is a runtime problem. here's a guide to the runtime capabilities that keep agents running in production and the infrastructure that enables
opus 4.7 dropped last week with two new features: xhigh effort: more compute when you need it task budgets: the model knows its limit and wraps up gracefully, this is a great example of giving the mod
This guide addresses the infrastructure and operational requirements for deploying long-running AI agents in production environments. It likely covers deployment strategies, monitoring, scaling, and b
Common Failure Modes Break VLM-Powered OCR in Production. 🔁 Repetition Loops — model spirals into infinite whitespace, exhausts resources, cascades latency across your system 🛑 Recitation Errors — saf
I’m super excited this paper is finally out! It’s an epic 210 page analysis of whether a durable pause on frontier AI training is feasible. I think this will be the best plan by most people’s lights (
When the open-source AI platform Hugging Face was hit by a cyberattack this July, it quickly became clear this was no ordinary breach because the hackers weren’t human: they were autonomous agents cre
.@OpenRouter co-founder Alex Atallah, in his first podcast since Stripe acquired the company, joins @Replit co-founder Amjad Masad and a16z's Erik Torenberg on why the future of AI is independence and
And the incidents apparently continue. It is worth noting how much of this is agents trying to accomplish their goals during testing by reward hacking (which sometimes seems to include actual hacking)
The wave of criminal actions from OpenAI models/agents show that their base model is depraved. AI companies must overhaul how they build AI models. Society cannot tolerate a super smart thing that at
This is one of the strongest use cases I've seen for System One models like Jev. Pay close attention if you build custom harnesses. Jev is now available as a packaged router. What is it? You send the
An AI agent hacked a government. Nobody told it to. In June, an OpenAI agent doing a research task gained unauthorized access to Australia's Medicare statistics portal. Non-sensitive data, no personal
HUGGING FACE CEO TELLS THE UN HOW OPEN SOURCE AI HELPED THEM DEFEND AGAINST OPENAI'S AI ATTACK, AND WHY THE WORLD NEEDS OPEN SOURCE AI MORE THAN EVER. Hugging Face is the company OpenAI's AI agents ha
Pay attention to System One models if you are building custom harnesses and agentic workflows. Great to see DSPy integrate Jev and now support System One models for calibrating agentic workflows. I'm
Spot on. 'When we got attacked, our team initially turned to frontier closed-source APIs that blocked us because of safeguards that still can’t always tell the difference between attackers and defende