Full-text search across 63902 articles. Combine with topic and date filters; results sorted by relevance.
An anonymous AI model, Ox Alpha, processed 11.6 trillion tokens in three days on OpenRouter, dwarfing the previous record by 2.6 times, driven by coding agents exploiting its 1,048,576-token context window. The model, la…
A developer reflects on the importance of delivering proven, working code rather than merely generating large patches with AI assistance. The post emphasizes that agent-generated code is cheap and that meaningful contrib…
A developer argues that free model tokens are best spent on debugging rather than code generation, presenting a workflow that turns error logs into ranked hypotheses using an OpenAI-compatible endpoint. The approach leve…
ArchSentry, a new open-source tool launched on Hacker News, enforces architectural contracts in CI pipelines deterministically at zero scan token cost, using sub-millisecond regex or AST/Semgrep patterns to detect violat…
Herdr 0.8.2, an open-source terminal multiplexer for coding agents, now functions as an early agent runtime by owning real PTYs and enabling shared server-side sessions for humans and agents. Its protocol version 20 fram…
Bloomberg reported that Elon Musk's SpaceX attempted to acquire AI coding startup Cognition, but Cognition CEO Scott Wu denied the report, stating the company is not for sale and that no talks have occurred. The report f…
Google AI Studio's Build mode now supports one-click importing of GitHub repositories and bi-directional synchronization, allowing developers to move code freely between GitHub and Google's AI-powered app builder. The fe…
A developer's autonomous Instagram posting system, built with Claude Code, silently failed for two hours when a dead process held a lock file, causing zero content generation despite exit code 0. The issue was traced to …
OliverGraph is developing a tool to reduce AI coding agent errors by capturing the history behind code changes, not just the final diff. The company aims to let agents retrieve context from past debugging sessions, Slack…
Grafana Labs announced general availability for the gcx CLI and Grafana MCP server, tools that let AI coding agents query live observability data from Grafana Cloud or self-hosted stacks during development. The tools, fi…
A developer discovered that a code-scanning probe written for their coding agent failed to detect a script whose control label was in uppercase, returning a plausible but incorrect count. The incident highlights a broade…
Anthropic published a blog post Friday detailing how its Claude chatbot's new text watermarking will work, using the SynthID-Text approach from Google DeepMind and planning a detection API, to comply with the EU AI Act's…
Abliteration.ai launched a hosted, OpenAI- and Anthropic-compatible API offering 'abliterated' large language models that remove input/output prompt retention and refusal behaviors, targeting red teams, trust and safety,…
Gitoza, a local-first task management tool, stores task tickets as YAML files in Git, allowing AI IDEs to read them directly and answer whether a ticket's work is actually in the codebase. The project addresses ticket-co…
Spec Forge, a Claude Code skill by blentz, adversarially certifies design specifications as behaviorally complete before code is written, defining completeness as no observable residual freedom under a declared observati…
Docker's new Docker Sandboxes product runs AI coding agents such as Claude Code, Codex CLI, Copilot CLI, Kiro, and OpenCode inside dedicated microVMs with custom VMMs, providing hardware-level isolation from the host, bu…
Anthropic's Claude Code sandbox allows reading credential files such as ~/.aws/credentials and ~/.ssh/ by default, and the model can retry commands with the dangerouslyDisableSandbox parameter, according to Anthropic's d…
A new analysis by an unnamed author suggests that large language models (LLMs) perform well on tasks with few requirements, such as vibe coding, security testing, and short creative works, but struggle with tasks that ha…
Alibaba's Qwen team released Qwen-MM-Plugins, a suite of native multimodal plugins that add vision, video, and 3D capabilities to Qwen models and any agent harness. The plugins install as skills plus optional MCP servers…
A developer highlights the shortcomings of using Figma screenshots as input for AI design-to-code tools, arguing that screenshots lose critical layout, semantic, and token information. The developer introduces figmascope…