AI agents
79 published articles
Enterprise AI Security
A synthetic dataset cut AI agent permission violations by 93%
A new arXiv paper proposes a dynamic permission architecture for enterprise AI agents, backed by a validated synthetic dataset. The approach reduced permission ceiling violations by 93% in testing, offering a proactive defense against over-privileged agents.
2026-07-30
Open Source Buzz Workspace Integrates Hermes Agent
Buzz and Hermes let ai agents share a channel with your coworkers
Buzz, an open-source workspace built on Nostr, now supports three ways to integrate Hermes, the autonomous agent from Nous Research. From a one-click desktop setup to a full gateway that keeps Hermes' memory and skills intact, the integration offers flexibility for different deployment needs.
2026-07-30
MemTensor's Memory Foundation Models
AI agents kept memory outside the model. Metis puts it inside
MemTensor's Metis is the first prototype of memory foundation models: history compressed into the backbone, read back through memory attention, updated in one forward pass. It challenges the vector-store era of agent memory. The limits are acknowledged but not yet quantified.
2026-07-30
AI Engineering
83 Claude Code tips just became a community wiki, and it reveals the new skill that matters
A community-maintained GitHub repo collects 83 tips for Claude Code, documenting the shift from vibe coding to agentic engineering. The resource reveals context budgeting as the new core skill.
2026-07-29
Artificial Intelligence
Google's five-step side hustle guide skips the one step that actually makes money
Google's Gemini side hustle guide is a solid planning checklist, but it skips customer acquisition, sales, and the grind of execution. The real bottleneck isn't planning, it's doing.
2026-07-23
Open Source
Your AI coding agents are running old playbooks. Nacos fixes that.
Managing Skills across multiple AI agents is a chaotic mess. Alibaba Cloud's Nacos Skill Sync offers a unified central registry and local sync modes, turning scattered prompts into shareable, version-controlled assets.
2026-07-23
Analyse comparative
14 microsecondes d'instanciation: PraisonAI peut-il vraiment détrôner LangChain?
PraisonAI promet 14μs d'instanciation et une simplicité redoutable face à LangChain et CrewAI. Mais sans cas d'usage en production, le pari reste à prouver. Analyse des forces et angles morts du framework.
2026-07-22
Gemini triple model launch lands July 21
Google ships Gemini models: cheaper Flash, faster Lite, and a vetted cyber variant
Google's three-model Gemini drop targets production AI with better token efficiency and lower latency. 3.6 Flash cuts token usage by 17% versus its predecessor. 3.5 Flash-Lite runs at 350 output tokens per second. A cyber-focused variant ships exclusively to vetted partners.
2026-07-22
Formal agent orchestration with stack-based execution
The 96% context cut that finally makes AI agent tool hierarchies work
A hierarchical, skill-based architecture for AI agents solves the bottleneck of flat tool registries. By organizing capabilities as a tree and using a stack-based execution loop, it scales sub-linearly with tool count while providing execution isolation and auditability for regulated domains like digital payments.
2026-07-21
Moonshot AI's scheduler for automated agent tasks
Your AI tasks now schedule themselves. Kimi Work ships a cron engine.
Kimi Work adds a built-in cron engine to the Kimi AI platform, enabling scheduled execution of LLM calls, data processing, and agent workflows. It targets users who need 24/7 automation without local infrastructure.
2026-07-21
Artificial Intelligence
The AI that taught itself to revise: four agents, one puzzle, no consensus
Researchers propose ARCANA, a multi-agent system that cracks ARC-AGI-2 puzzles by splitting reasoning into four steps and closing the loop with reflective feedback. Each failure teaches the next attempt. The architecture improves reasoning efficiency, but questions about generalization and compute cost remain.
2026-07-19
New benchmark exposes agent reasoning limits
AI agents hit 49% on a test human experts pass at 95%. More compute won't fix it.
Alibaba's HSCodeComp benchmark reveals the gap: top agents hit 49.4% accuracy vs. human experts' 95% on tariff classification. The bottleneck is structural, more compute doesn't help.
2026-07-19