Full-text search across 31918 articles. Combine with topic and date filters; results sorted by relevance.
A veteran programmer who sold a company to Google reports achieving 40-50x coding productivity with AI over 16 weeks, measured in lines of code, and says the boost is increasing. The programmer, who previously wrote up t…
The GCC project's contribution guide expresses skepticism toward AI-generated code, citing that LLMs often produce plausible but incorrect logic that maintainers lack bandwidth to debug. The guide permits AI use only for…
Hamza, a proxy for Claude Code and Codex, masks secrets and personal data before prompts reach Anthropic or OpenAI, reducing exposure of credentials and customer records that coding agents may inadvertently include in re…
A developer found that AI-generated Flutter code consistently makes seven structural mistakes, including bad state management, missing tests, and hardcoded values. One example involved an agent using a subprocess call to…
Anthropic shipped Claude Code, its agentic CLI coding assistant, on Bun instead of Node.js, a choice that highlights tradeoffs in startup latency, bundling strategy, and native module compatibility for TypeScript-heavy d…
Code review remains the hardest part of software development even as LLMs make code creation cheaper, according to a developer reflecting on open-source maintenance. The author notes that while LLMs can help find problem…
A developer built a fully autonomous AI system that reads a GitHub issue, understands a codebase, writes working code, runs tests, gets code reviewed by another AI, and opens a pull request—all without human intervention…
Anthropic engineer Thariq Shihipar reported on July 24 that the team deleted more than 80% of Claude Code's system prompt for Claude 5 generation models with no loss on coding evals, calling the change "unhobbling." The …
SessionWatcher, a macOS menu bar app developed by Soren Starck, monitors token consumption and usage limits for AI coding tools including Claude Code, Codex, and Cursor, displaying real-time usage, limits, and cost estim…
Anthropic reduced the system prompt powering its Claude Code coding agent by approximately 80%, from roughly 800 tokens to 164, with no performance loss and potential improvement, announced engineer Thariq Shihipar on Ju…
A developer released Curated Claude Code, a small agent harness for Claude Code that prioritizes safety and self-evolution over bloated skill collections. The system features an intake gate (/vet) that blocks unverified …
Code mode, which prompts LLMs to generate executable code instead of structured tool-call signatures, can help smaller models perform tool calling more reliably without additional training, according to a developer analy…
A developer benchmarked Claude Code agent skills against a placebo and a no-instruction baseline, finding that half of the tested skills failed to outperform the placebo. The most-starred 'Karpathy Guidelines' skill beat…
Alibaba's Qwen2.5-Coder 32B outperforms Meta's Llama 3.1 70B in code generation tasks, delivering 35 tokens per second versus 12 tokens per second and a 2.1-second time-to-first-token compared to 4.8 seconds, according t…
Vals AI launched Vals-Smith on July 22nd, a tool that converts a GitHub repository's merged pull requests into a private benchmark for testing coding models and agents. The product addresses the problem that a model lead…
Microsoft released Visual Studio Code 1.129 on July 15, introducing a dedicated agent host that runs agent harnesses like Claude, Copilot, and Codex via the Agent Host Protocol, enabling sessions to persist across multip…
Poolside released Laguna S 2.1 on July 21st, an open-weight 118-billion-parameter Mixture-of-Experts coding model designed to run long-running software agents on user-controlled hardware. The model activates 8 billion pa…
A developer built Full-Stack Quality Skill, a reusable AI coding skill for full-stack audits, architecture guidance, long-term project memory, and CI quality gates. The skill enables AI agents like Cursor, Codex, and Cla…
A developer at Vancine, a third-party API aggregation platform, tested Kimi K3 on a single coding task using OpenCode, finding it cost $0.19 to complete a leap-year fix in 84 seconds. The run used 28,707 tokens and passe…
A developer shared in a DevOps discussion that an AI code reviewer on a Node and React project repeated style complaints, commented on outdated code, and missed a null check that caused a crash, requiring full human revi…