Retiring the Fork: Why ForgeClaw Became a Private System
ForgeClaw is no longer being presented as a public WebForge product. It remains under active private use, but the public position has changed.
// april_2026_notice
ForgeClaw is actively being used for private proprietary purposes and undergoing constant evolution. The change is public positioning, not internal abandonment.
What Happened
When Greyforge first adopted OpenClaw, upstream was not yet moving at the pace our own work required. We needed routing behavior, memory discipline, specialist roles, and workflow assumptions that fit our own operators immediately, not eventually.
So we forked. That was the right move for that phase. ForgeClaw gave us the freedom to move without waiting and the freedom to harden ideas under real workload pressure instead of just theorizing about them.
The result was useful because it was specific. ForgeClaw worked for us because it was shaped around our own habits, our own infrastructure, our own timing, and our own tolerance for complexity.
Why It Mattered
The fork proved an important point. A custom operator system can outrun a general open project for a while when the problem is narrow and the feedback loop is immediate.
That phase has limits. Everyone building in this space ends up solving for their own reality. Their own hardware. Their own habits. Their own risk tolerance. Their own idea of what should be local, what should be automated, and what should stay under direct human control.
What we built works for us. That does not mean it should be presented as though it naturally fits everyone else.
The Shift
OpenClaw did not stand still. It advanced quickly, and it is still advancing quickly. Credit is due to Peter Steinberger and the wider open source community for that pace.
That changes the economics of maintaining a separate full-stack fork. We cannot outpace the entire world at scale while also carrying every audit, upgrade, merge, and compatibility burden of an independent runtime.
Greyforge also has other active projects that need time, work, and attention. A permanent fork only makes sense when the maintenance burden remains decisively worth it. That is no longer true in the same way.
The Architectural Response
The response is not abandonment. It is consolidation. Greyforge is adapting more day-to-day work back into the main OpenClaw repo where that makes sense, while moving differentiated work into modular tools, skills, policies, and private operating layers that can evolve without dragging a whole parallel runtime behind them.
That line is more honest and easier to maintain than pretending every useful internal mechanism should remain a public product surface.
What Changes Publicly
ForgeClaw is being deprecated from the WebForge product catalog. The historical chronicles remain because they are real records of what was built and why, but they should now be read as lineage, not as a standing public sales pitch.
- ForgeClaw is actively used privately.
- ForgeClaw is proprietary.
- ForgeClaw is under constant evolution.
- ForgeClaw is not being marketed as a general product.
- Supporting utilities may still ship publicly through OpenForge when they stand on their own.
What Comes Next
The next phase is disciplined integration. Greyforge will keep building the systems it needs, keep documenting the lessons that survive contact with reality, and keep publishing standalone utilities when those tools are genuinely useful beyond Greyforge itself.
ForgeClaw is not dead. It simply belongs where it was always most valuable: inside the forge, under active use, changing constantly, and serving the work that actually depends on it.