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.
The RL update improves hard math and code scores to match the base Qwen3.8-27B while maintaining 0.67x thinking token usage on agent tasks.
Released under the MIT license, the agent operates in observe mode by default and requires human approval for auto-merges.
Mistral released the weights for its 1T MoE model, Mistral Large 4, demonstrating token efficiency.
The research prototype listens to GitHub and Feishu to open work items, execute tasks in isolated sessions, and report results via pull requests.
tstack, released by Theo Browne, is a set of 9 skills that enable coding agents to handle the process up to PR merging.
The built-in provider supports Supervised and Auto modes using existing fx models and skills.
Vercel has released durable-harness, which allows coding agent turns to continue via logs.
StepFun released Step Code, derived from Pi, with built-in sub-agent and web deployment capabilities.
The new dataset contains 19,652 records from real coding sessions, while the renamed Scaffold Chat dataset now has 3.45 million examples after removing tool calling.
Corgi improves terminal management efficiency through agent state tracking and background session persistence.
The Apache-2.0 licensed tool runs the pi coding agent in-process across isolated workspaces, available for macOS, Windows, and Linux.
The free beta tool compiles repositories into a queryable map to reduce source code reading by 2-6x compared to mainstream agents.
The open-weight model reduces thinking tokens to roughly half of the base model's usage while maintaining valid tool calls in agent loops.
A tool that returns relevant files and source code based on behavior-oriented queries in unfamiliar codebases.
The update introduces a single background service process, a native JavaScript interpreter for tool orchestration, and asynchronous background subagents.
The 0.8B Jeff model determines whether to think on each turn, reducing Qwen3.8-27B's task time by 32%.
The new Jeff-Code agent uses a small decision model to pre-fetch information and manage thinking levels, reducing average task time to 0.68x of the baseline with no significant drop in accuracy.
Supports 12 backends including Claude Code and offers 14 official flows
A local server that delivers existing Markdown/Obsidian documents to coding agents via CLI, HTTP, and MCP without modification.
The Emacs package agent-shell, based on the ACP protocol, has been released, enabling control of 19 coding agents through a single UI.
OpenAI has released GPT-6 Sol and Luna via AI Gateway, offering lower pricing than GPT-6 Astra.
Vercel AI Gateway now supports Z.ai's GLM 5.3 FlashX model, which enables ultra-fast inference.
AWS has streamlined the new account creation process and introduced $100 free credits along with coding agent integration features.
Hugging Face has released 'Tau', a lightweight coding agent inspired by Pi.