Open Source Radar — October 8: The Agent Wranglers

Today's trending list is less about new frameworks and more about the awkward middle of agent development: supervising fleets of coding agents, auditing what they touch, and keeping them fed on long tasks. Four fresh projects, all verified on GitHub.

cmux (Swift, ~28,000 stars) is an open-source macOS terminal built specifically for running AI coding agents — Ghostty-based, with vertical tabs, per-agent notifications, and everything exposed through a CLI and Unix socket so agents can drive the terminal themselves. It's climbing daily trending because a terminal designed around parallel agent sessions is suddenly a normal thing to want. Use it as the cockpit when you run more than one coding agent at a time.

security-audit-skill (JavaScript) is Cloudflare's open-source coding-agent skill that turns your agent into a multi-phase security auditor: reconnaissance, coverage-led hunting, validation, then machine-readable findings that get independently checked before they ship. Cloudflare says this skill seeded the internal vulnerability-discovery harness it now runs fleet-wide, so this is the single-repo starting point of a production system. Grab it if you've ever wondered whether your agent can actually review code for security holes rather than just style.

hallmonitor (Go) is a read-only dashboard for your coding-agent fleet: every Claude Code and Codex session — on your laptop and your SSH boxes — shown live in a terminal board, menu bar, or right around the MacBook notch, with who's working, who's stuck, and what each one is doing. It launched on Product Hunt today, and the privacy stance is the selling point: it never sends prompts or approves anything, it just watches. Reach for it when you've lost track of which terminal that agent was in.

pmtui (Rust) tackles the longer-horizon version of the same problem: a tmux-backed supervisor that keeps one persistent Claude or Codex session per project nudged toward a goal you set, auto-handles the ordinary choices, and only pulls you in when a decision genuinely needs a human. The insight is that agent loops drift on long tasks, so pmtui keeps the loop outside the agent and rebuilds the prompt from your goal plus the agent's own last status each time. It hit Product Hunt today with its first releases this week — a bet on supervision layers over agent runtimes.

Worth watching this week.

Supervising agents is becoming its own tooling category — which of these would survive a real week of your workflows? Tell us in the comments. Sources: cmux (GitHub) · security-audit-skill (GitHub) · hallmonitor (GitHub) · pmtui (GitHub)

Akhetonics says its all-optical CPU reaches a customer in 2026

Light-based computing keeps promising more than it delivers — but one Munich startup has just put a date on its bet, and the interview laying it out is doing the rounds on Hacker News this week. Akhetonics says it will deploy its first commercial machine with a major customer by the end of 2026, with several more planned for 2027. The company, founded by Michael Kissner and Leonardo Del Bino, is building a computer where data enters as light, is switched as light, and circulates through memory

The Week in AI — October 5–11, 2026

Every big claim this week turned out to rest on fine print more interesting than the headline: revenue only the company reporting it can define, safety tests sandboxed while the product keeps the web, and a Pentagon phase-out nobody would confirm until reporters kept asking. The week's top 5 1. OpenAI's revenue was $20 billion below the numbers everyone quoted — and the gap was definitional. The Financial Times reported Thursday that OpenAI's annualized revenue runs roughly $20 billion unde

Drone strike shuts a third Yandex data center, taking YandexGPT offline

Russia's largest tech company is learning what the AI era's infrastructure war looks like from the receiving end — three data centers in four days, and with them much of the cloud layer Russian businesses run on. A Ukrainian drone strike knocked out Yandex's data center in Vladimir early Sunday morning, the third of the company's facilities hit since October 8. The site — reported at roughly 50 MW and designed for about 2,880 server racks — stopped operating completely after the attack, Yandex

Agent teams cost up to 5x more, barely score higher

The multi-agent hype train hit a benchmark this weekend — and the grid and the trucking regulators had quiet weeks of their own. Vals AI put agent teams head-to-head with single agents on its Vibe Code Bench, and the teams cost between 1.8 and 5.1 times more for almost no extra quality. The evals company ran GPT-6 Sol and Claude Opus 5.5 solo and in teams across 50 apps at two reasoning efforts; out of four comparisons, only one was statistically significant — Sol at medium effort, where the t

Read more

Akhetonics says its all-optical CPU reaches a customer in 2026

The Week in AI — October 5–11, 2026

Drone strike shuts a third Yandex data center, taking YandexGPT offline

Agent teams cost up to 5x more, barely score higher