Never think solo.
The bottleneck moved.
For two years we’ve been racing to make individuals faster. Copilot, Cursor, Claude Code, Devin — every tool tuned to shave another second off your loop. By most measures it worked. An engineer with a good agent ships more code in a Tuesday than they used to ship in a week.
But the team isn’t shipping faster. Not by the same multiple. Often not at all.
The bottleneck moved. Individual velocity isn’t the constraint anymore. Team coherence is.
What goes missing in agent-augmented work.
Maggie Appleton put a name to it in One Developer, Two Dozen Agents, Zero Alignment. When everyone has their own private agent thread, three things go missing across the team:
- Shared context. Nobody knows what anybody else asked their agent yesterday.
- Shared priority. Each thread has its own implicit todo list. None of them line up.
- Shared verification. The artifact in your head is not the artifact in mine.
The work happens, but it forks. Pull requests stop being proposals and start being archaeology.
What unsolo does about it.
unsolo is a multiplayer Claude session. One chat thread, shared. One todo list, shared. One live preview, shared. Three primitives mapped one-to-one to the three things that go missing.
You open a room. You drop the URL into Slack. The PM, the designer, the engineer, the UX writer all land in the same room at the same time. The chat is shared. The todos are shared. The preview is shared. Decisions get pinned at the top so the reasoning stays traceable.
You don’t pull anyone’s branch. You don’t paste transcripts at each other. You think together — with Claude, in one room.
Who it’s for.
Product teams that work in the early-shape phase: figuring out the wedge, sketching the wireframe, drafting the empty state, naming the thing. The part of the work that needs more than one head.
Not for solo coding. Not for async standups. Not for the parts of work where being alone is the feature, not the bug.
The promise.
Never think solo.
When the work needs more than one head, give it more than one head — in the same room, on the same artifact, at the same time.