Full-text search across 59718 articles. Combine with topic and date filters; results sorted by relevance.
A developer built Anansi, an open-source personal web memory that lets coding agents search and retrieve saved references via a read-focused MCP server. The browser extension captures saves from X, Reddit, GitHub, and de…
Developer Avinash Amudala released Weftgate v0.1.1, a local static verification gate that checks environment variables, Python and Node imports, and FastAPI/Starlette route handlers in agent-written code before it ships.…
Anthropic shipped the `claude plugin eval` command in Claude Code v2.1.269, a tool that scores plugins against test cases and reports a delta metric comparing results with and without the plugin loaded. Anthropic's docum…
A developer demonstrated that AI-generated tests can degrade coding-agent repair performance, citing an ExecCritic preprint by Leitian Tao and colleagues in which weaker Qwen Test agent feedback dropped Qwen-3.5-35B-A3B'…
OpenAI released GPT-6 Astra, a model focused on computer use, coding, professional workflows, science, and cybersecurity, initially available to a limited set of organizations and rolling out to ChatGPT Plus, Pro, Busine…
Novence launched a static-site hosting service that lets coding agents deploy a site end to end via a single command, `npx novence deploy ./dist`, which uploads files, publishes, and prints the live URL at `https://{suff…
Microsoft's Edge team said on Tuesday that the rapid adoption of AI-assisted coding has led to a surge in browser extension submissions, straining its review pipeline and increasing turnaround times. To cope, the team ha…
Andy Geek released Blackholes, an open-source macOS app for organizing coding agents, Git repositories, and terminals, supporting Claude, Codex, Gemini, and OpenCode. Built with Rust and React, it allows projects to cont…
Felan, an open-source coding agent focused on efficiency, reports a 36.6% reduction in API costs across six controlled comparisons, with individual cost reductions ranging from 5.2% to 66.0% while maintaining quality. Th…
OpenAI reported that its coding agents now handle over 50% of internal research tasks, reducing experiment cycle times by up to 70%, according to a company blog post. The shift enables faster iteration on model architect…
WorkBraid, a new developer tool, visualizes architecture diffs for agent-proposed code changes, storing state as Git-backed JSON and rendering before/after diagrams for human review. It offers a CLI and MCP server, enabl…
A developer has released chamnan v1.22.1, a local pre-processing plugin for Claude Code that sanitizes, budgets, and organizes data before it reaches the AI model. Built in pure Python with over 3,500 automated tests, th…
Coding agents are evolving as harnesses get absorbed into models, with specialized agents rising to fill gaps where baseline models lack data, according to an industry analysis. The commoditization of models shifts value…
South Korean security firm Genians Inc. reported that North Korea-linked hacking group Kimsuky used the open-source AI coding agent opencode to create decoy documents for malware distribution, marking the first such dete…
Simon Willison demonstrated that Blender can be controlled by coding agents on macOS by installing the full Mac application from blender.org and prompting ChatGPT Codex to render a scene of a pelican riding a bicycle, th…
A developer warns that AI-generated tests are not independent checks of code but rather reflections of the same assumptions the AI used to write the implementation, leading to correlated errors that coverage metrics fail…
Aidcrew, a new open-source terminal tool from developer Antonio Ciccia, orchestrates a team of coding agents that each run on different AI models and providers, with every job isolated in its own git worktree. The tool, …
Jern Cloud launched a governed cloud coding agent that operates under a repository-pinned policy file, with sessions running on isolated machines, hard token caps, and encrypted evidence, delivering pull requests for hum…
A developer built Superagent, a Mac app that gives Claude Code and other AI agents a real computer environment to work in, rather than a chat window. The app consists of an Electron desktop app, a SwiftUI iOS companion, …
Developer Christine Seeman released cseeman/claude-journal, a Claude Code plugin that writes structured session summaries to a Notion work journal. The plugin creates one callout per topic with a bold headline, metadata …