An MCP server exposes a data tool's capabilities to AI agents through the open Model Context Protocol, letting agents such as Claude Code, Cursor, Codex, and ChatGPT call a tool's functions directly i…
Agnostic-ai, an open-source tool from developer Chemaclass, generates native configuration files for 25 AI coding CLI targets from a single Markdown-and-YAML source stored in the user's repository, wi…
A newly released paper applies Sutton's Bitter Lesson to data agents, arguing that as large language models improve, general coding agents will absorb the hand-engineered scaffolding that human-design…
A researcher launched NetHackers, an open project to find a program that can reliably win NetHack 3.6.6, after roughly 1.5 months of development. The project lets participants hand-write a bot, have C…
A community-built tracking website called Codex Reset checks public news sources every 5 minutes to monitor global, banked, and personal Codex reset signals, according to the site. The tool logged rec…
A developer has proposed a role-based model routing system for AI coding sessions, assigning tasks to Navigator, Builder, Diagnostician, or Surgeon roles based on the cost of feedback rather than proj…
VillageSQL released vsql-mcp, an extension that runs a Model Context Protocol server inside the MySQL database server process rather than as a separate sidecar, serving MCP over Streamable HTTP from a…
OpenAI released ChatGPT Voice "on-the-go" on Wednesday, extending its upgraded GPT-Live voice model to the web and mobile versions of its agentic tools, including ChatGPT Work and Codex. Pro and Plus …
OpenAI announced on Wednesday that it is bringing voice-based agentic features to the ChatGPT mobile app, letting users trigger workflows such as drafting documents, drafting emails, and summarizing S…
Anthropic's own testing shows Claude Code's auto mode misses 11% of harmful actions, and novel prompt injection techniques can reliably execute malware when auto mode is on, according to a Latacora se…
Agent-Native released Agent-Native Plans, a free, open-source tool that lets coding agents produce visual documents for planning and review, built around two skills called Visual Plan and Visual Recap…
Solo, a new developer tool created by Aaron Francis, combines terminal management with deep AI agent integration and a built-in MCP server that gives coding agents access to a user's tasks and termina…
OpenAI announced a new Community Trainer Program on September 23, 2026, marking the second anniversary of OpenAI Academy by certifying local instructors to teach AI skills in their own communities. Th…
GPT-6 Astra, run through Codex at medium reasoning effort, completed 100% of the DrivingBench cone course in 5:22, the only model to finish among four frontier models given control of a Toyota Corolla…
Local AI Weekly issue #3 argues that generous Claude and Codex subscription pricing amounts to a "token rug pull," warning that credit giveaways, degrading models, faster credit burn and eventual pric…
Agnostic-ai released version 0.66.0, adding a settings spec `effort` key that sets the default reasoning effort for Claude Code, Copilot, Codex, and Factory under each tool's own key (#1069, #1089). T…
Sharkly.ai is a free platform for managing multiple AI agents and human teammates within a single project workflow, according to a discussion of tools for running 3–5 coding agents in parallel. The po…
Developer Yijun Yu released jev-rs, a Rust crate that turns any GGUF-served LLM into a probability-scoring engine for typed questions — yes/no (noul), choice, and score — returning probabilities inste…
Parity engineers ran a self-hosted inference trial of the open-weight DeepSeek V4 Flash model that handled more than 144,000 requests and almost 12.9 billion tokens from 25 engineers between 16 August…
Developer Viktor Petersson released Agent IAP, an open-source identity-aware proxy that pulls credentials from 1Password so AI agents receive narrow, time-boxed tokens instead of long-lived API keys. …