Openframe is the directory's open-source entry. It is a small, maintainer-driven project that publishes a reference implementation of a basic agentic operating system — the kind of system the Bulletin's coverage spends a lot of time describing, but as code rather than as architecture diagrams. The Bulletin tracks Openframe because the project's existence is, in our view, the strongest available evidence that Web4 is becoming a real category and not only a market narrative.

The project is maintained by three independent engineers. Avi Shoenfeld is in Tel Aviv and has the largest commit footprint. Mira Tanaka is in Osaka and handles documentation, release management, and the larger architectural decisions. The third maintainer publishes under the contributor handle "colt-2008" and has not disclosed a public identity; the Bulletin has not pushed for more information than the project itself has chosen to publish.

The reference implementation is deliberately minimal. Openframe ships a small core — a job-distribution loop, a specialist-agent abstraction, a structured handoff format, an audit log, and a pluggable UI layer that defaults to a card-based surface. It does not ship integrations with any specific cloud provider. It does not ship a billing layer. It does not try to be production-ready for a real business. The README is explicit that the project is a reference implementation, intended to be read, forked, and learned from rather than deployed.

The Bulletin's interest in Openframe is partly that it is one of the only open-source artifacts in the directory and partly that it is one of the only public documents that lets a reader interact with the operating-system pattern at the code level. Several of the Bulletin's essays — particularly Aksoy's cornerstone piece on what Web4 is — point readers to Openframe as a recommended way to ground the architectural claims in something they can run on a laptop.

The project's editorial value is also a function of what it has refused to do. Openframe has, repeatedly, declined to accept sponsorship from any of the larger platforms the directory tracks. The project has stayed deliberately small, deliberately minimal, and deliberately uncommercial. Tanaka has been the most public about this, in a written exchange with the Bulletin earlier this year, and her stated reason — that a reference implementation loses its reference value the moment it starts optimizing for a specific platform's strategic interest — is the kind of position the Bulletin's editorial team has tried to make room for.

What is harder to evaluate, and what the Bulletin notes here, is the project's long-term durability. Three-maintainer open-source projects historically face structural risk around bus-factor, contributor recruitment, and the gradual erosion of attention that affects most volunteer-led work. Openframe is at the 0.6 release as of the most recent commit window. The cadence has been steady. The contributor base has been growing slowly but consistently. We expect to track the project's durability with the same patience we extend to similar maintainer-driven projects in adjacent categories.

For Bulletin readers who want to engage with Web4 at the code level rather than only at the level of architecture diagrams and product launches, Openframe is the directory's recommended starting point. The repository is the project's home, and the maintainers welcome serious technical contributions even though they have been firm about turning down commercial pressure.