A Codex CLI login hang at the "Finish signing in via your browser" prompt was traced to an 18-hour-old `codex app-server --managed-daemon` process (PID 21226) holding the OAuth callback port localhost…
JoyDemo cut bugs by about 95% by having its team use AI to make product changes directly on the production host rather than building in a separate development environment, according to the company's a…
A developer published an open SKILL.md template that configures AI assistants to act as a patient senior Python instructor rather than an instant answer generator. The template, tested in Lumo AI and …
DigitalOcean launched Managed Agents in public preview this week, packaging sandbox, inference, and data as a single runtime for AI agents on top of its existing Droplets, databases, and networking. T…
A developer released AGENTBOX, a zero-dependency, fully local flight recorder that wraps any AI agent or command and produces a hash-chained JSONL session file of every tool call, shell command, file …
Specific Labs' Real-SWE benchmark, which evaluates frontier models on private enterprise codebases across 640 rollouts, found that Fable 5.1 running in Claude Code achieved the highest first-try resol…
Apple shipped a native Model Context Protocol server in Safari 27 on September 17, making Safari the first major consumer browser to expose live DOM, network requests, console output, and screenshots …
Researchers reported a Codex sandbox escape in which untrusted code generated by the AI agent read authentication tokens from a shared process memory region in the Node.js utility bundled with Codex D…
Developer totocaster released Copy Editor, an open-source MIT-licensed local prose editor that runs in a browser with a server on the user's own computer, where writing, notes, revisions, and export w…
A headless Codex CLI provider added to mager-bench ran GPT-5.6 Sol through all 13 coding challenges for an average score of 9.0/10, with Doom at 8.7, Slots at 8.3, and async-fetch the low point at 6.3…
A developer's analysis of coding agent architecture shows that swapping only the harness around a fixed model — keeping weights, tasks and context window constant — raised SWE-bench Verified bug-fixin…
A private coding-agent evaluation called Bakeoff recorded 277 runs across 10 harness-and-model combinations, spanning 10 models and 3 harnesses (Claude Code, Codex CLI, and pi) at an approximate cost …
A developer benchmarked Polars 2.0.0-rc.1 against a cached-parser workaround on a Mac mini M4 Pro, finding the new release processed one million date strings in 0.13 s — about 25× faster than the lru_…
Crest, a Mac notch app, now surfaces Claude Code and OpenAI Codex CLI permission prompts in the MacBook notch so developers can tap Allow or Deny without returning to the terminal. Crest uses Claude C…
A developer detailed a year-long setup for running coding agents like Claude Code and Codex CLI on real products, built around a single 32-core, 128 GB Linux box reached over Tailscale and managed wit…
A developer built jev-blindspot, a lightweight tool for Claude Code and Codex CLI that uses a local hook and background daemon to detect "blind spots" in prompts before they are submitted. A classifie…
Researcher Oren Yomtov of Accomplish AI disclosed two OpenAI Codex sandbox escape vulnerabilities, Heapjack and Overpatch, that were reported on August 12, patched within eight days, and publicly disc…
DigitalOcean opened its Managed Agents product to general availability following a private preview, letting teams deploy agent harnesses such as OpenCode and Codex CLI, connect agents to more than 16,…
A developer built ContextVeil, an open-source, local tool that strips selected secret values from coding agent tool output before it reaches the model's context. It uses deterministic, exact matching …
A developer built ChibiRigKit, an open-source MIT-licensed tool that converts a single character illustration into an animated, editable 2D rig. The pipeline uses Codex CLI for AI-assisted preparation…