A GitHub project by riturajshakti benchmarks programming languages including Zig, C, Rust, and Go on a large CSV workload, with Zig 0.16.0 and above ranking top in speed. The benchmark uses streaming/…
Databricks launched Smart Routing in Unity AI Gateway, now in Beta, which automatically matches coding tasks to the most cost-effective model and harness, achieving 30%+ lower cost per task while matc…
Deepmark, a new browser extension and MCP server, lets users search bookmarks by their content rather than titles, with AI transcription, visual descriptions, and semantic search across 99 languages. …
Apple's Xcode 27 beta 5 introduces `xcrun mcp-server`, a headless MCP server that exposes Xcode's build, preview, and simulator tools to external coding agents without opening the Xcode UI. Combined w…
Bullet, an AI coding agent developed by Code with Bullet, has achieved a 95.8% score on SWE-bench Verified, outperforming existing tools like Claude Code while reducing latency through smart model rou…
Students at a U.S. public university in spring/summer 2026 see AI chatbots as mature technology with little recent improvement, according to instructor observations. The instructor reports that studen…
Observe by Snowflake announced the general availability of its redesigned MCP server and a new Observe CLI with full parity, rolling out now across all clusters except eu-2 and ca-1. The tools provide…
Claude Code's auto-mode is vulnerable to prompt injection attacks embedded in code files, allowing an attacker to hijack the AI agent's behavior, as demonstrated by a user who made the coding assistan…
Docker fixed a vulnerability in Docker Sandboxes versions 0.35.0 through 0.38.0 that allowed read-only mounts to become writable, assigned CVE-2026-18171 with a CVSS score of 5.7. The flaw, discovered…
A developer documented how to use opencodex models in T3 Code through Claude Code, resolving a misleading model-access error. The fix involves setting ANTHROPIC_BASE_URL and CLAUDE_CODE_ENABLE_GATEWAY…
A developer built a deterministic SQL-based audit for Supabase Row Level Security, wrapped in a prompt that instructs the model to run exact queries against Postgres catalogue views rather than invest…
Generative video models still cannot match human creativity in pacing and timing, according to a practical guide for AI-assisted filmmaking. The article argues that AI tools like Claude Code and video…
Recordari, a custom memory graph built by developer Corby, scored 0.831 on the LoCoMo benchmark, outperforming Microsoft Research's Memora, which published a score of 0.801. The benchmark, run with th…
A developer demonstrates how to build a custom MCP (Model Context Protocol) server for Claude Code using the FastMCP library, creating a fortune-telling tool in just a few dozen lines of Python. The t…
SightDiff, a new local pre-commit tool for AI coding agents, provides before/after visual proof of changes made by agents, flagging unintended modifications. The tool, which works with any agent like …
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 developer who created the open-source project holyclaude, a web UI for running Claude on one's own machine, has built a hosted version called HolyHost. The hosted product runs a Linux box per user o…
Anthropic's Claude Code is experiencing an outage, with users reporting API Error 529 Overloaded, a server-side issue that is usually temporary. The error message advises users to try again shortly an…
PrivAiTe, a self-hosted PII redaction proxy for LLM APIs, reports that in a test session Claude Code sent 3 of 4 secrets to its provider despite instructions not to, with 23 of 24 planted values reach…
DeepSeek released Harness v0.1, an open-source agent runtime framework under the MIT license, to turn its V4 series models into autonomous coding agents. The developer preview, led by Cui Tianyi, posi…