Full-text search across 60653 articles. Combine with topic and date filters; results sorted by relevance.
A developer argues that AI code review tools often prioritize finding every possible issue over providing useful feedback, leading to a poor developer experience. The developer, who is building a tool called Codzee, emph…
A developer has created Port, a tiny macOS window that displays Claude Code agent session states in the blank space to the left of the dock, providing a persistent view and notifications when sessions need input or are c…
Researchers Matteo Cardoni and Sam Leroux introduced a training technique for Predictive Coding Networks (PCNs) that pairs a Generative PCN with a support Encoding PCN, trained in parallel to match neural activations wit…
A developer introduced a copy-paste completion protocol for AI coding agents, defining what agents must show before claiming software work is done. The protocol requires final reports to include four fields: DONE, PROOF,…
CodeYam CEO and cofounder Nadia Eldeib narrowed her personal website redesign to four design variants in Part 4 of her Substack series, iterating with CodeYam and Claude Code. Eldeib incorporated design feedback from Dan…
Developers report that Anthropic's Claude Code is randomly consuming large portions of weekly token limits in brief periods, with one user using 9% of their limit in 10 minutes after a reset, according to Hacker News thr…
A developer built a token-optimization stack for coding agents to reduce API costs, but killed the project after discovering that the savings numbers were misleading and the rigorous benchmark would cost over $1,200 in A…
TheGitAI, a model-agnostic coding agent for the terminal, has been launched on Hacker News, offering a default ask-before-edit mode, auto-accept, and plan mode, with journaled edits and checkpoint-based reversibility. Th…
A non-technical founder built learnfrom.co, a platform for creating, hosting, and selling online courses, using AI to generate code, and it now serves 100+ creators and businesses with real payments. The founder is seeki…
A developer replaced an LLM call in a letter generator with 200 lines of deterministic template code, citing fixed output shape, legal-adjacent risk, and zero marginal cost as key reasons. The pure function approach elim…
AWS Professional Services has published production data on a multi-agent system that compresses infrastructure-as-code development from weeks to minutes. The system chains four specialized agents—discovery, IaC generatio…
Anthropic and Google Cloud are hosting a webinar on September 1, 2026, at 12:00 PM PT to discuss controlling costs and measuring ROI for Claude Code on Google Cloud. The session, led by Roy Arsan of Anthropic and Ivan Na…
Meetless has launched a runtime context system that coordinates parallel coding agents, detecting stale or conflicting instructions and surfacing only changes requiring human judgment for approval. The platform provides …
A developer, Ammar Najjar, shares his local toolkit for making AI coding agents more efficient, focusing on controlling information flow rather than adding capabilities. He documents tools like Headroom, RTK, Caveman, Se…
Coding agents suffer from context rot long before their context windows fill, degrading judgment and treating prior errors as ground truth, according to a developer advocating the 'Ralph loop' pattern. The Ralph loop, na…
Henka, an open-source multi-tenant Model Context Protocol (MCP) server, enables AI agents to perform structured, semantics-aware code refactorings across Java, Rust, and TypeScript/JavaScript projects using each language…
A Hacker News user reports that coding agents they launched in parallel sometimes complete tasks that were already abandoned, submitting patches with passing tests for plans the user no longer wants. The user asks how ot…
A developer who built a coding agent from scratch to compare it against the minimal harness pi discovered that pi's minimalism is about disciplined scope, not line count. The developer found that pi's core remains four t…
An engineer at Softjourn nearly installed a malicious package recommended by an AI agent, but company policy requiring verification on GitHub prevented the installation. Separately, security researchers at Wiz identified…
Endor Labs has launched an AI-powered Static Application Security Testing (SAST) tool for C code that detects more vulnerabilities than traditional scanners without requiring a software build. By combining deterministic …