EElay osetrov56m ago · 4 min readYour vendor review asks where the data lives. It doesn't ask what the agent can do.Most vendor security reviews were written for tools that store data. AI agents also act: they read content nobody vetted, hold credentials, and call tools. A questionnaire that stops at "where is datasetrov.hashnode.dev
GGermán Giménez1h ago · 5 min readWhy we stopped routing every AI task to the most expensive modelI'm Germán, CTO and co-founder of Napsix (NapsixB2B), a small team in Mendoza, Argentina. We build what we call an agentic workspace: a place where a company's people, its tools and AI agents work toggermangimenez.hashnode.dev
EElay osetrov6h ago · 4 min readYour AI agent just did something it shouldn't. Here's the first hour.Picture the message in your team chat: "Why did the support bot email a customer's invoice to an address nobody recognizes?" If your agent can call tools, this will happen to you in some form. It mighsetrov.hashnode.dev
EElay osetrov3h ago · 3 min readCan user A's question pull user B's documents? A tenant-isolation checklist for RAGMost RAG data leaks don't need a clever attacker. They need one missing filter. A shared vector index, a retrieval call that forgot the tenant_id, a semantic cache keyed only by the question text, andsetrov.hashnode.dev
LLaxman Joshi21h ago · 3 min readEvernote Meets Cognee - Without Stale MemoriesIllustrative scenario below. I did not have access to a live Evernote account while building this, so everything described as "tested" was tested offline against fake Evernote clients. The problem Yea2LAcognee-evernote.hashnode.dev
HHaithem Slimi1d ago · 30 min readTeaching a Small Language Model to Triage Incidents: A Beginner’s Unsloth and QLoRA TutorialMost fine-tuning tutorials stop when the training loss decreases. I wanted to answer a more useful operational question: can a small language model learn to turn an unstructured incident alert into a 24KNnightthoughts.me
TThe Vibe Check5h ago · 4 min readMistral Large 4 weights, Jules shuts down, Pro API creditsOriginally published in The Vibe Check, a free daily 90-second email on what changed in AI agents and coding tools. Issue #14, Oct 30, 2026. 🛠️ Get the Vibe Check Agent Skills Pack ($12): 12 ready-tothevibecheck.hashnode.dev
AAshen Jayawardena5h ago · 23 min readThe Tester's Guide to RAG Evaluation, Part 2: Building the EvalsIn Part 1 you met the RAG system: retrieval finds passages in a corpus, generation writes an answer from them, and each half fails in its own way. You spiked the system to see how it really behaves, aashenasserts.hashnode.dev
MMagenta Waterwheel5h ago · 7 min readConverting PDFs to Clean Markdown for RAG and LLM PipelinesIf you've built a RAG pipeline that ingests real-world documents instead of clean .txt files, you've hit this problem: PDFs don't contain text the way you'd want them to. They contain layout instructimagentawaterwheel.hashnode.dev
Jjaberadam200119h ago · 9 min readWhy Your AI Agent Calls the Wrong Tool (and How to Test Tool Routing)TL;DR: An AI agent doesn't look tools up by name. On every turn it reads every tool description and picks the closest match for the request. That makes tool selection a relative decision, so adding on36KGDsimplyexplained.hashnode.dev
