Tech News
What xAI's Grok Build CLI Actually Sends to xAI
xAI's Grok Build CLI uploads entire repository including git history to xAI servers, regardless of what the agent reads, raising privacy concerns.
vllm-project/vllm v0.25.0
vLLM v0.25.0 makes Model Runner V2 the default for dense models, removes PagedAttention, and speeds up the Transformers backend to match native vLLM performance.
The new GPT-5.6 family: Luna, Terra, Sol
OpenAI's latest flagship model hit general availability this morning , and comes in three sizes: Luna, Terra, and Sol (from smallest to largest). The new models are priced per 1M input/output tokens as Luna $1/$6, Ter.
Build a semantic layer for agentic AI on AWS with Stardog and Amazon Bedrock AgentCore
In this post we show how to build a semantic layer on AWS using Stardog’s Semantic AI Application over Amazon Aurora and Amazon Redshift, and how to run a Strands Agents agent on Amazon Bedrock AgentCore that queries.
How Deutsche Telekom is rewiring telecommunications with AI
How Deutsche Telekom is becoming an AI-native telco with OpenAI-transforming customer service, employee workflows, network operations, and the future of voice.
GitHub Repos
bytedance/deer-flow
An open-source long-horizon SuperAgent harness that researches, codes, and creates.
# Python# agent# agenticruvnet/ruflo
🌊 The leading agent meta-harness.
# TypeScript# agentic ai# agentic frameworkcan1357/oh-my-pi
⌥ AI Coding agent for the terminal — hash-anchored edits, optimized tool harness, LSP, Python, browser, subagents, and more
# TypeScript# ai agent# ai coding agentResearch Papers
Cognitive-structured Multimodal Agent for Multimodal Understanding, Generation, and Editing
This paper presents a multimodal agent that uses episodic visual memory and structured retrieval to handle long conversations, outperforming larger models on a synthetic benchmark while being faster.
UniClawBench: A Universal Benchmark for Proactive Agents on Real-World Tasks
UniClawBench is a new benchmark for evaluating AI agents on real-world tasks, focusing on five key capabilities like skill usage and exploration, using live Docker environments and multi-turn feedback.
Early to Share, Late to Save: Synchronisation-Driven Communication Gating in Bandwidth-Constrained Cooperative VLN
A lightweight supervised gate is proposed that labels communication-critical steps post-hoc from navigation failures, avoiding the high variance of REINFORCE and establishing a new communication regime for bandwidth-limited embodied agents.