Categories

Coding agents

An abstract coding agent session timeline flows into a secure audit collector

GitHub exposes Copilot agent sessions for enterprise audit

Abstract code blocks pass through build and deploy checkpoints before meeting a harder behavior-validation maze

ScarfBench shows coding agents still struggle with Java migrations

Developer tasks pass through a policy gate into several abstract model routing lanes with a cost meter below

GitHub turns Copilot model choice into a governed routing layer

A coding agent workflow loops through traces, grading, failure clusters, and approved fixes

Google gives coding agents an eval flywheel instead of another prompt tweak

Google's new quality-flywheel skill lets coding agents run structured agent evaluations with independent grading and production-trace loops.

The AI Feed Desk

By The AI Feed Desk

A streamlined agent workspace with a smaller model core coordinating tools beside a larger reference model

Claude Sonnet 5 turns Anthropic's default model into an agent model

Anthropic made Sonnet 5 the default for Free and Pro users while positioning it as a lower-cost agentic model close to Opus 4.8.

The AI Feed Desk

By The AI Feed Desk

A coding-agent session timeline branches into tool calls, cache blocks, and latency markers

TraceLab turns real Codex and Claude Code sessions into serving data

The University of Washington TraceLab release studies coding agents as production workloads, with public traces across sessions, tool calls, tokens, cache behavior, and latency.

The AI Feed Desk

By The AI Feed Desk

An AI-generated pull request passes through a software review gate with quality checkpoints

Cognition's FrontierCode asks whether AI code would survive review

FrontierCode evaluates coding agents on mergeability, code quality, scope, tests, and maintainer judgment instead of only functional correctness.

The AI Feed Desk

By The AI Feed Desk

A bright solar model core is held behind a transparent safety gate

OpenAI previews GPT-5.6 Sol behind a limited-access safety gate

OpenAI's official feed says GPT-5.6 Sol improves coding, science, and cybersecurity capability, but the public body was blocked to plain fetch, so the responsible read starts with access and verification.

The AI Feed Desk

By The AI Feed Desk

A testing gauge compares a clean tool path with a longer tangled debugging path

Hugging Face measures whether tools are agent-friendly

Hugging Face's agent-focused benchmark tests whether software changes help coding agents finish tasks with fewer errors, tokens, and detours.

The AI Feed Desk

By The AI Feed Desk

A coding benchmark workbench with agent paths, repository blocks, cost meters, and verification checkmarks

DeepSWE makes coding-agent rankings a cost question

DeepSWE's June 20 leaderboard update separates frontier coding agents by pass rate, cost, output tokens, and agent steps across long-horizon software tasks.

The AI Feed Desk

By The AI Feed Desk

A small coding model module routes glowing code paths to laptops, phones, IDE panels, and command-line surfaces

MAI-Code-1-Flash moves across GitHub Copilot before enterprise access

GitHub says Microsoft's small coding model is expanding across Copilot CLI, app, chat, IDE, mobile, and Xcode surfaces before Business and Enterprise rollout.

The AI Feed Desk

By The AI Feed Desk

A long code ribbon passing through a model core and into a completed software system

Z.ai releases GLM-5.2 for long-horizon coding work

Z.ai's GLM-5.2 pairs a 1-million-token context pitch with long-horizon coding benchmarks, public docs, API pricing, and an MIT-licensed Hugging Face model card.

The AI Feed Desk

By The AI Feed Desk

A developer automation pipeline where task cards flow through sandbox and review checkpoints

GitHub Agentic Workflows brings agents into Actions

GitHub Agentic Workflows is in public preview, letting teams define natural-language automations that compile into Actions workflows with runner and safety controls.

The AI Feed Desk

By The AI Feed Desk

An abstract coding workspace showing agent task lanes moving code blocks between editor panels

GitHub moves Copilot toward agent-first development

GitHub's VS Code Copilot changelog and Microsoft's Visual Studio 2026 notes show Copilot becoming a managed agent workflow across sessions, models, terminals, and C++ modernization.

The AI Feed Desk

By The AI Feed Desk

A dark developer terminal with multiple parallel agent session lanes flowing into a central dashboard

xAI puts Grok into Warp and adds a dashboard for coding agents

xAI's June 15 updates give Grok and X Premium subscribers access to Grok models inside Warp and add a dashboard for supervising parallel Grok Build sessions.

The AI Feed Desk

By The AI Feed Desk

Generated editorial illustration of an AI coding interface connected to orbital compute infrastructure

SpaceX moves to buy Cursor for $60B

SpaceX's planned $60B Cursor acquisition turns AI coding agents into a strategic asset tied to compute, developer workflow, and xAI infrastructure.

The AI Feed Desk

By The AI Feed Desk

Editorial illustration of Grok Build installing plugins for deployments, debugging, browser control, and cloud workflows

xAI launches Grok Build Plugin Marketplace

xAI's Grok Build Plugin Marketplace adds installable bundles for skills, commands, agents, hooks, MCP servers, and LSPs inside its coding agent.

The AI Feed Desk

By The AI Feed Desk

Generated editorial image of a command-line console connected to model repository cards and agent nodes

Hugging Face redesigns the hf CLI for coding agents

Hugging Face says Claude Code and Codex are the largest coding-agent cohorts on the Hub, and its redesigned hf CLI cuts token waste and command probing.

The AI Feed Desk

By The AI Feed Desk