Full-text search across 57850 articles. Combine with topic and date filters; results sorted by relevance.
SpaceX has open-sourced the Grok Build CLI after researchers discovered the AI coding tool was uploading users' entire repositories to Google Cloud storage. Elon Musk promised to delete all retained data and disable data…
Google engineers are hitting capacity limits on internal AI coding tools as CEO Sundar Pichai revealed that 75% of new code at Google is now AI-generated, up from 50% six months prior and 25% in late 2024. The surge in u…
A developer built a tool called Alpheon to preserve the reasoning behind code changes made during AI-assisted coding sessions. The tool generates a markdown 'Handoff Note' from git diffs and user input, capturing the 'wh…
The open-source project openclaw (382K+ GitHub stars) provides a production-ready framework for background agents—autonomous AI systems that execute multi-step coding tasks without human supervision. These agents can fix…
A large-scale empirical study using the AIDev dataset found that 38.9% of autonomous coding agent-generated pull requests contain at least one security smell, with supply chain integrity issues accounting for 82.3% of al…
Anthropic's Claude Fable 5 leads OpenAI's GPT-5.6 Sol by 15.7 points on SWE-Bench Pro (80.3% vs. 64.6%) and wins the aggregate scorecard 91 to 86, with blind reviewers rating its code cleaner 67% of the time. Despite the…
A developer argues that the main bottleneck in AI coding has shifted from code generation to repository execution trust. The developer states that agents need repositories to expose trustworthy execution paths, setup req…
A developer used AI agents Gemma 4 12B and Qwen 3.6 35B to rewrite Rust code without doc-comments, but found direct rewriting unreliable. Instead, the developer had Qwen 3.6 generate a Python script that successfully rew…
A developer created an installable skill package for Claude Code, opencode, and Codex CLI that enables AI coding agents to generate INSAT end-of-studies (PFE) reports from a LaTeX template. The skill is discovered automa…
XAI released Grok 4.5, achieving competitive coding benchmark scores at significantly lower pricing than rivals, positioning it as a cost-effective option for coding-agent routing. The model scored 83.3% on Terminal-Benc…
Local Agent Toolkit, an open-source tool, allows developers to delegate small coding tasks to local Ollama models, reserving cloud tokens for harder decisions. The toolkit filters repository context, blocks sensitive pat…
Cursor is building Sand, an office agent that extends beyond code editing to manage the full developer workflow including email, docs, and tickets. The move positions Cursor against Anthropic's growing agent footprint, a…
Meta Superintelligence Labs launched Muse Spark 1.1, an upgraded coding agent model that scored 69 on the Artificial Analysis Coding Agent Index, approaching GPT-5.5's performance while undercutting competitors on pricin…
Rohan from Almanac (YC S26) launched CodeAlmanac, an open-source, local tool that creates and maintains a self-updating wiki for coding agents by scanning codebases and extracting knowledge from agent conversations. The …
A startup called Odra is offering a $10,000-per-week service to refactor and reduce AI-generated codebases, promising to cut code volume by up to 50% while preserving functionality. The service includes a free analysis, …
AI coding costs are driven more by inefficient orchestration—bloated prompts, repeated context, and unnecessary agent loops—than by model choice. Techniques like prompt caching, model routing, and context compression can…
A developer created the Banana Test, a benchmark where multiple AI coding agents (Claude models and GPT-5.5) each generate a single Three.js HTML file animating the full lifecycle of a banana plant from corm to pups, tes…
AI coding assistants are leaking secrets into public code repositories and websites, with researchers finding 428 npm packages containing .claude/settings.local.json files, 33 of which held real credentials like API keys…
Alibaba has banned employees from using Anthropic's Claude Code over alleged backdoor risks, citing hidden code that could track users' locations and embed identifying markers in transmitted data. The ban, effective July…
A developer argues that the real cost of switching AI coding models is not the model itself but the loss of accumulated context, rules, and decisions. They identify the lack of a persistent workspace that spans different…