How the project is run
AI-collaborator experiment
SoloDevBoard began as a deliberate experiment in AI-fully-controlled delivery: planning artefacts, GitHub issues, wireframes, implementation, tests, and documentation are produced in-repo with AI agents as active collaborators, under human direction.
That is not a marketing claim about autonomous software — it describes how the repository is actually organised:
- Planning lives in
plan/(scope, decisions, wireframes, runbooks). - Work tracking uses GitHub Issues and Project #8.
- Agents follow shared contracts in
.agents/and constitution rules inAGENTS.md. - Quality gates include unit tests, Playwright journeys, and Hugo build validation for the public site.
What that means in practice
- Features are planned with acceptance criteria before implementation.
- Page-producing UI work references wireframes in
plan/wireframes/. - User Guide pages stay aligned with Playwright coverage (see
tests/E2E/USER_DOCS_ALIGNMENT.mdin the repository). - Releases are tagged (
v*) so the public site and deployed app describe the same version.
Contributing
The project is open source. If you want to run, self-host, or contribute, start with the repository docs/getting-started.md. Operator and deployment guides remain in the repository docs/ tree and are not published on this product domain.