Tech News
Show HN: Clawk – Give coding agents a disposable Linux VM, not your laptop
Clawk provides a disposable Linux VM for coding agents, enhancing security by isolating agent operations from the host machine.
datasette code-frequency chart on GitHub
Out of curiosity I decided to see if I could find a useful illustration of the impact of coding agents and Opus 4.5 class models on my own output. The best I've found so far is this GitHub chart of frequency of code.
Implement on-behalf-of token exchange for multi-tenant agents with Amazon Bedrock AgentCore Gateway
Building multi-tenant agents with Amazon Bedrock AgentCore and Apply fine-grained access control with Bedrock AgentCore Gateway interceptors establish the conceptual foundation for on-behalf-of (OBO) token exchange.
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.
Interrail: 6,379Km and 13 Countries over 7 weeks
A traveler recounts a 7-week, 13-country Interrail journey covering 6,379 km, with community discussion on the changing nature of train travel and pass value.
GitHub Repos
OpenHands/OpenHands
🙌 OpenHands: AI-Driven Development
# Python# agent# artificial intelligencedataelement/bisheng
BISHENG is an open LLM devops platform for next generation Enterprise AI applications.
# TypeScript# agent# aijrswab/axe
A lightweight cli for running single-purpose AI agents.
# Go# ai agents# automationResearch Papers
RAGU: A Multi-Step GraphRAG Engine with a Compact Domain-Adapted LLM
RAGU is an open-source GraphRAG engine that improves knowledge graph construction by separating entity extraction from consolidation, using a compact 7B LLM that outperforms larger models on graph tasks.
HyperSafe: Inference-Time Safety Recovery for Fine-Tuned Language Models
HyperSafe uses a hypernetwork to generate a small safety classifier for each fine-tuned LLM, routing harmful prompts to refusal without modifying the model weights.
Evidence-Backed Video Question Answering
This paper introduces E-VQA, a new video QA task that requires models to output both an answer and precise spatio-temporal evidence (temporal segments and object masks). It also provides a benchmark and a large-scale instruction dataset to train models for this grounded...