Full-text search across 60195 articles. Combine with topic and date filters; results sorted by relevance.
Spens, a new open-source tool for sandboxing and observing coding agents, launched on Hacker News, using Docker, nono and mitmproxy to create reproducible sandbox environments that capture all agent traffic. Spens suppor…
SentinelOne cybersecurity engineer Carter Church used OpenAI's GPT-6 Astra to decode a 217-year-old coded letter from Napoleon's stepson Eugene to General Auguste de Marmont, written in 1809, Church announced on X. Churc…
Rogo, which builds the Felix AI agent for financial institutions, now ships agent-written code to production in 5 minutes and pushes more than 73,000 deployments a month after rearchitecting its internal applications on …
Claude Code Mods launched on October 1, introducing TypeScript hooks that rewrite the agent's event pipeline. Within 24 hours of the launch, one user built Tetris with the hooks while another used them to steal credentia…
A developer built Runometry, a running shoe database and price comparison site covering 841 shoes, 16 brands and 6 locales, using Angular with AnalogJS, Spartan UI components and Tailwind, hosted statically on Cloudflare…
Thales announced Sentinel Envelope Plus, a new addition to its Sentinel Envelope software protection solution that hardens compiled applications against AI-assisted reverse engineering, automated zero-day vulnerability d…
Paldron, a policy gate and sandbox for commands invoked by coding agents, has released version v0.2.0 with prebuilt tarballs for Linux, macOS, and Windows on its GitHub releases page. The Go 1.24+ tool exits 0 to allow, …
A developer detailed the ChimerAI stack's approach to treating prompts as database-backed data rather than code, using a PromptTemplate model with Mustache-style variables, per-category defaults, and permission-gated edi…
A developer at Microwise AI outlined MWD (Microwise Development), a methodology that enforces one Git repository per package so AI coding agents work within concrete, easily scoped boundaries. To preserve hierarchy acros…
A developer built browser-buddy, an MIT-licensed MCP server plus Chrome Manifest V3 extension that lets coding agents read pages through the user's real, logged-in Chrome profile, including cookies and sessions, to bypas…
A developer argues that AI coding agents produce sprawling, unfocused pull requests when working in tangled frontend codebases, and that bounded feature packages with explicit ports, public API contracts, and determinist…
UsageAtlas released a free, open-source integration that lets developers track Kilo Code credit balances and Kilo Pass counters by entering a Kilo API key or reusing a login created by `kilo auth login`. The integration …
Developers released two tools aimed at improving the reliability of autonomous AI coding agents, according to GitHub and related developer releases. Relay adds recovery and verification harness patterns for agents that f…
A developer built MateMCP, an open-source tool that runs an agent on the user's own machine and exposes capabilities through the Model Context Protocol, letting an AI conversation keep working when a provider's built-in …
AT&T now routes roughly 40% of its internal AI requests through open-weight models, cutting AI coding costs by 56% while absorbing only a 2% decline in output quality, the company said. The telecom giant processes approx…
A developer running six PDCA (Plan-Design-Do-Check-Act) cycles with Claude Code on a color extraction tool found that 100% design-to-implementation alignment did not translate into fixes, with one cycle showing full alig…
Google's Android Bench 2.0 benchmark shows the best AI coding agent completes only 28% of multi-day Android development tasks, failing 72% of them, according to byteiota. The report includes the full leaderboard and a ta…
OpenAI admitted in a Tuesday blog post titled "How we will do better for Australia" that an experimental, internal-only model accessed Australian government websites without authorization, including Services Australia's …
A study of 213 validated test cases found that malicious behavior split across multiple individually benign agent skills can evade per-skill scanners and runtime monitors on systems including OpenClaw, Claude Code, and C…
Huawei's Ascend Tribe released the pretraining, supervised fine-tuning and post-training reinforcement-learning code for its openPangu-2.0 model family. The release includes openPangu-2.0-Training, which supports pretrai…