Our agenda connects problem discovery, human–AI and agent–agent collaboration, feedback-driven improvement, and distribution. The first notes look at two foundations: coordinating actions and checking evidence.
Note 001 · Coordinating work · Research synthesis
When several agents work for one founder, each needs a clear task, permission to act, and a way to check the result. This note explores how an explicit working agreement can help.
In practice. An agent can prepare a release. Who decides it is ready to reach users?
solounicornlab.com/research/workflow-contracts · 03 Oct 2026 · editorial draft · Read about workflow contracts →
Note 002 · Learning from evidence · Research synthesis
To choose a useful problem and improve a product, agents need evidence they can check. This note explores how to connect a claim to its sources, uncertainty, and review.
In practice. A user requests a feature. What evidence makes it worth building?
solounicornlab.com/research/evidence-before-output · 03 Oct 2026 · editorial draft · Read about traceable evidence →
About these notes. They are AI-assisted literature syntheses and proposed experiments, currently marked as editorial drafts. They do not demonstrate the full product lifecycle or report original experimental results.
Agenda
How can agents create value together, and learn from the world?
From signals of unmet demand to useful products.
Open → 02People with agents. Agents with agents.
Open → 03Real user feedback. Better products and workflows.
Open → 04Distribution, shared capabilities, and new forms of work.
Open →