The Stack

Open Source Radar — August 22: swarms, workspaces & observability

Editorial

Editorial

— 2 min read
Share
Open Source Radar — August 22: swarms, workspaces & observability

Today's board is the agent operating layer: a meta-harness for orchestrating teams of agents, Apache's new local-first agent workspace, and an analytics platform that wants to be every AI team's eyes on what agents are doing. Three projects, all fresh on daily trending, none of them repeats.

ruflo (TypeScript, ~68k stars) — An agent meta-harness from the developer behind claude-flow, built to deploy intelligent multi-agent swarms and coordinate autonomous workflows from one place. It pairs adaptive memory with RAG integration so a team of agents shares context instead of each starting from a blank slate, and it plugs natively into Claude Code, Codex, and Hermes rather than squaring you into a new harness. It's the biggest star mover on today's board because teams keep hitting the ceiling of single-agent setups — running one agent is manageable, but running several that swap state cleanly between a "hive mind" is the harder problem it actually targets. If you're building agent teams and tired of stitching them together by hand, this is the most finished meta-orchestrator on the daily list.

Maka (TypeScript, ~2k stars) — Apache Maka, an incubating project under the Apache umbrella, is a local-first agent workspace that records every model message, tool call, tool result, and permission decision as an append-only log — the log is the runtime, and your sessions and recovery are just projections over it. It ships as a desktop app (currently an early build for macOS Apple Silicon), plus a terminal TUI and a non-interactive CLI, all behind controlled permissions so the agent can inspect projects, run tools, and produce artifacts without free rein. Local-first by default and Foundation-backed, it's early but credible for anyone who wants agent work that is recorded, recoverable, and reviewable rather than a chat transcript you hope you can rewind.

PostHog (~38k stars) — The open-source product analytics suite has sharpened its pitch to building "self-driving products": it layers AI observability — tracing your LLM app's generations, latency, and cost — on top of the usual session replay, feature flags, experiments, and error tracking, and you can steer it from your editor via MCP or from Slack. Its self-driving mode reads the signals in your product data (errors, rage clicks, failed queries) and turns them into researched reports and pull requests for a human to review and merge. It lands on daily trending because teams building AI features now need eyes on their agents the way they once needed analytics on their web users — and PostHog is betting that observability becomes the control layer for that, with a generous free tier.

Worth watching this week: the agent stack keeps getting thicker — orchestration frameworks, record-everything workspaces, and observability platforms are converging on the same idea, that agents need infrastructure, not just a stronger model.

Would you let a "self-driving" analytics tool file pull requests on production code, or keep a human in every loop? Tell us in the comments.

Sources: ruflo (GitHub) · Maka (GitHub) · PostHog (GitHub)

Read more

Nvidia-backed Firmus pulls its $30 billion IPO after demand collapses

Nvidia-backed Firmus pulls its $30 billion IPO after demand collapses

The biggest AI-infrastructure listing of the year didn't get delayed — it got pulled entirely, after two price cuts in two days found no takers. Nvidia-backed data-center operator Firmus withdrew its Australian IPO on Friday, abandoning what would have been the country's second-largest listing ever after slashing the offer price twice in 48 hours. The company had been marketing shares at A$11 each to raise about $5 billion at a valuation near $30 billion — roughly three times the $10.5 billion

AI inference is redrawing the storage hierarchy

AI inference is redrawing the storage hierarchy

The money has followed GPUs for years, but at GMIF 2026 in Shenzhen the argument was about everything behind them — and the numbers backing it are getting hard to ignore. China's 140 trillion daily token calls are forcing a redesign of the storage stack. Now in its fifth year, the Global Memory Innovation Forum gathered storage makers, analysts and chip designers around one theme: inference, not training, is now the workload that dictates hardware. China's daily token calls had already passed

Cloudflare launches Clef-omni, cuts Clef-flash price by 58%

Cloudflare launches Clef-omni, cuts Clef-flash price by 58%

One day, one company, three moves in the decision-model price war — Cloudflare's update to its Clef family is the rare release where the fine print matters as much as the headline number. Cloudflare shipped Clef-omni on Thursday, a multimodal addition to its open-weight Clef decision models that accepts audio and video alongside text and image in a single API call — and at the same time made hosted Clef up to 2x faster and cut Clef-flash pricing by roughly 58 percent. Clef-omni takes WAV or MP

Meta turned down Amodei's personal plea for compute

Meta turned down Amodei's personal plea for compute

The chip hunt is back in the headlines — and today's edition runs from a boardroom ask at the very top of the AI industry to the economics of putting robot drivers in truck cabs. Dario Amodei personally approached Meta earlier this year to source more compute for Anthropic, and Meta declined — according to the Wall Street Journal. The report lands inside the Journal's larger piece on how desperately the industry is hunting for computing power, and it is the kind of detail that only surfaces wh