Full-text search across 65014 articles. Combine with topic and date filters; results sorted by relevance.
GitHub made dynamic workflows available in public preview across Copilot CLI, the GitHub Copilot app, and the GitHub Copilot SDK, letting developers define multi-agent orchestration in code. The feature runs on all Copil…
A developer added DeepSeek as the tenth agent in the OpenSpec-UI project by writing a single new file, deepseek-acp.ts, that implements the shared AgentAdapter interface over the Agent Client Protocol, with no changes ne…
Accomplish detailed the engineering behind its coding-agent sandbox, which runs an Ubuntu guest OS inside a full virtual machine on macOS so that the agent harness, downloaded code, guest user and guest root are all trea…
Vostride AI released agent-qa, a QA agent that exercises real web and mobile applications alongside coding agents and returns evidence to the team, under the Functional Source License (FSL-1.1-ALv2) as a Fair Source proj…
Fireworks AI released Ember-1, a post-trained version of Moonshot AI's open-weight Kimi K3 that delivers Kimi K3's quality with about 40% fewer tokens, available only through the Fireworks serverless API as a Research Pr…
Fireworks AI released Ember-1 on September 22, 2026, a fine-tuned variant of Moonshot AI's Kimi K3 that cut reasoning token consumption 71.3% and total token spend 39% in live production A/B tests on coding workloads, wi…
WorldScript Studio, an open-source writing studio, is architected so its AI features are optional rather than load-bearing: all AI capabilities route through a single provider seam (Gemini, OpenAI, OpenRouter, Anthropic,…
A developer at KandiForge describes hitting the "last mile" bottleneck in agentic software development, where hundreds of AI agents can generate millions of lines of code across 20 repositories but the final 10-20% of ha…
A developer working on account-deletion logic found that a coding agent's generated tests passed while the implementation deleted accounts before verifying eligibility, because the tests asserted only on HTTP status code…
Amazon Web Services extended its Model Context Protocol (MCP) server with new AI agent skills for AWS End User Messaging and Amazon SES, letting developers configure messaging workflows through plain-language commands in…
A developer built The Estian Tattler, a tabloid-style news site for a RimWorld colony in which a Claude reporter writes gossip drawn from a real save file and a plain TypeScript fact-checker rejects any claim naming a co…
Rails published a landing page and benchmark suite pitching the framework as a good fit for coding agents, arguing that convention over configuration, Ruby's token density, and decades of public Rails code make agent-gen…
A developer built an AI SRE system that dispatches a team of four specialist agents — covering code, releases, infrastructure, and metrics — plus a synthesizer and a skeptic to diagnose the first hour of on-call incident…
A developer used Claude Opus 5.5 inside Claude Code to build a self-documenting music video entirely in code, starting from the open-source PDoomVideo project. The pipeline renders every frame as a pure function of time …
Developer Yijun Yu released jev-rs, a Rust crate that turns any GGUF-served LLM into a probability-scoring engine for typed questions — yes/no (noul), choice, and score — returning probabilities instead of generated text…
LocalStack demonstrated that giving Claude Code access to its LocalStack MCP server let the AI agent deploy a 36-resource AWS CDK application, detect data loss after a DynamoDB table rename wiped all records, and restore…
Stanford computer scientist James Zou and colleagues published Paper2Agent in Nature on September 16, an open-source framework that converts academic papers and their codebases into runnable AI agents. Fed AlphaGenome do…
A new benchmark called GameReplica, presented in arXiv paper 2609.22308v1, evaluates whether vision-language coding agents can replicate games as runnable programs using only screenshots and an action interface, with no …
Google released the managed agent identifier antigravity-preview-09-2026 on September 17 and set a hard shutdown date of October 5, 2026 for antigravity-preview-05-2026, after which any application, scheduled workflow, o…
Xiaomi released and open-sourced the MiMo-V2.6 series, two natively omnimodal models for coding, visual tasks and computer use, with MiMo-V2.6-Pro scoring 46.32 on the Artificial Analysis Intelligence Index v4.3, which X…