AI Briefing#coding-agent
The latest AI and developer news about #coding-agent, with the original source and a short summary.
The latest AI and developer news about #coding-agent, with the original source and a short summary.
It's an ACP adapter that treats local coding agents like Vercel AI SDK models.
Warp has unveiled a terminal-based agentic development environment along with a new open-source repository.
During large-scale GLM-5 serving, a KV Cache race condition fix and LayerSplit optimization were disclosed.
Poolside has unveiled the Laguna XS.2 and Laguna M.1 models, optimized for coding agent performance.
A Tree-sitter-based local memory layer dramatically reduces coding agent token usage.
The open-source coding agent Dirac scored 65.2% on Terminal-Bench-2 and 8/8.
OpenAI announced it is stopping SWE-bench Verified reporting and shifting to new evaluation axes.
Anthropic disclosed 3 causes behind the Claude Code quality degradation.
Combining summarization, voting, and re-refinement boosted the performance of long-horizon coding agents.
Retrieval selection boosted coding agent performance across 9 software tasks.
Anthropic's revenue has surged 3x, but a shortage of compute resources is prompting discussions of price increases to manage demand.
The little-coder and Qwen3.6-35B-A3B combo scored 78.7% on Aider Polyglot.
Maestro v1.6.3 expands specialized agents and Qwen Code support.
Even when shown their budget, coding agents couldn't cut their own spending.
Shared a setup for running Qwen3.6-35B-A3B with llama-server on an RTX 4060 8GB VRAM, along with a thinking budget pitfall.
With the same Qwen3.5-9B Q4, simply changing the scaffold pushed Aider performance from 19.11% to 45.56%.
Lazyagent, a terminal TUI tool that lets you monitor the behavior of multiple AI coding agents like Claude Code at a glance, has been released.
Qwen3.6-27B, an open-source sparse MoE model optimized for agentic coding, has been released.
Qwen3.6-35B-A3B has been released, achieving large-model-class coding and multimodal performance with just 3 billion active parameters.
Vercel Labs' open-source reference app for background coding agents.
This is a real-usage warning that GLM 5.1 overwrote existing md and cron files without confirmation.
A coding agent model, Qwen3.6-35B-A3B, has been released for anyone to use.
The open-source CLI coding agent v0.2 for 8k-context LLMs shows a diff before modifying files.