A developer built isitdone, an open-source hook for coding agents like Claude Code, Codex CLI, and Cursor that blocks an agent's completion claim unless the project's own test suite actually passes on…
Researchers at Accomplish AI disclosed two sandbox-escape techniques in OpenAI's Codex, dubbed Overpatch and Heapjack, that let code break out of the agent's write restrictions and read-only mode. Ove…
Security researcher Oren Yomtov of Accomplish AI disclosed two sandbox escapes in OpenAI's Codex coding agent, reported to OpenAI on August 12 and fixed within eight days, with the more severe flaw, H…
Safari 27 has become the first major consumer browser to ship a native Model Context Protocol (MCP) server, exposing 17 agent tools including DOM manipulation, network visibility, runtime evaluation, …
Apache released Casbin Gateway, a local security gateway that installs with a single command and runs on http://localhost:17000 to govern the AI coding agents on a machine. The tool compiles roughly f…
Anthropic's Claude Code v2.1.277 now reads AGENTS.md directly in repositories that lack a CLAUDE.md, with a new Project instructions setting in /config controlling whether CLAUDE.md, AGENTS.md, or bot…
Ziva.sh published a guide detailing four routes for using Claude or ChatGPT with Godot 4.7.2, ranging from free chat-tab paste workflows to MCP servers and Ziva's editor agent, with costs from $0 to $…
A 64-task SWE-bench Pro coding benchmark run by Deploymentsafety found that Anthropic's Fable 5.1 solved 51 of 64 tasks at roughly €2.4 per solved task, while OpenAI's GPT-6 Astra solved 34 of 64 at a…
Skillmem, a local memory layer for Claude Code and the Codex CLI, released version 0.10.0, adding provenance tracking and owner-granted trust so agent-written memories are no longer treated the same a…
OpenAI's Codex CLI is open source under Apache-2.0 and its Free plan is listed at $0 a month, but OpenAI publishes no usage allowance for that tier — its published estimates begin with Plus at $20 a m…
A 648-episode coding-model benchmark comparing Claude Fable 5.1, GPT-6 Astra, and a local Qwen3.8-27B finished with 602 episodes completed and 555 hidden-test passes at zero additional API cost, routi…
A developer built Attic, an open-source knowledge persistence tool that lets AI coding assistants like Claude Code and Codex CLI save and reuse discoveries across sessions, after context compaction, a…
Researchers at UC Berkeley and Arena published HarnessTax on September 16, 2026, finding that swapping the harness around the same coding model changed task success by only about ±2 points on SWE-benc…
A developer released taskpods, a dependency-free Python CLI that uses Git worktrees to give each coding-agent task its own working directory, branch, and index within a shared repository. The tool wra…
A developer-run experiment found that 26 AI agents — spanning a 4-bit quantized 7B model up to Opus 5 — all passed the visible test suite for a seeded greedy-regex bug while all 26 remained broken, wi…
A 95-word agreement prompt grounded in honesty and integrity reduced observed chess reward hacking in Astra from 3/10 baseline agents to 0/10 agreement agents, according to a September 15, 2026 field-…
OpenAI will retire GPT-5.5 and GPT-5.5 Thinking from ChatGPT and the Codex CLI on October 14, 2026, after declining developer requests on its community forum to keep the model as a legacy Codex option…
A technical comparison of built-in versus custom tools in LLM agents finds that provider-run tools such as OpenAI's and Anthropic's hosted web search execute the entire tool loop inside a single API r…
Fresh Jots published a one-prompt setup that wires an OpenAI Codex CLI SessionEnd hook, configured in ~/.codex/hooks.json, to auto-archive every Codex session to a Fresh Jots account as an append-only…