Documentation topic
Delivery & Operations
Core guides
Start with the architecture
These guides carry the main argument for this topic. The full library below goes deeper into individual implementation decisions.
Our deployment workflow
Why meaningful changes start on branches, become real preview deployments, pass production builds and audits, and only then reach the public site.
Read the guideWhy preview and production are treated as different systems
A preview URL is not a miniature production site. It has a different audience, indexing policy, review purpose, and risk profile, so we make the environment distinction explicit in code.
Read the guideSecurity headers and site hardening
How we reduce avoidable attack surface on public websites with secure defaults, deliberate headers, simpler dependencies, and environment-aware configuration.
Read the guideOwnership, handoff, and long-term maintenance
How repository ownership, portable content, documented integrations, and predictable deployment reduce switching cost and make long-term maintenance practical.
Read the guideWhy regression prevention is part of the engineering, not just QA
Why regression prevention belongs in the engineering system itself, with repeatable checks that catch breakage before visual QA has to rediscover it.
Read the guide10 guides
All Delivery & Operations documentation
Why meaningful changes start on branches, become real preview deployments, pass production builds and audits, and only then reach the public site.
What a staging environment is for, what it is not for, and why real branch deployments improve design review, mobile QA, indexing control, and release confidence.
How we reduce avoidable attack surface on public websites with secure defaults, deliberate headers, simpler dependencies, and environment-aware configuration.
How semantic HTML, keyboard behavior, focus, readable contrast, motion preferences, and component testing shape our accessibility implementation.
How repository ownership, portable content, documented integrations, and predictable deployment reduce switching cost and make long-term maintenance practical.
Why many marketing forms validate and hand leads to a dedicated provider instead of turning the public website into a second customer database.
Why repository history matters during handoff, debugging, rollback, and long-term ownership, not just while the original development team is still involved.
A preview URL is not a miniature production site. It has a different audience, indexing policy, review purpose, and risk profile, so we make the environment distinction explicit in code.
Why regression prevention belongs in the engineering system itself, with repeatable checks that catch breakage before visual QA has to rediscover it.
A practical handoff checklist for custom websites: source code, repository access, domains, deployment, content, analytics, integrations, and the information needed to leave cleanly.
From explanation to proof
Where this connects to the work
This category documents the work that keeps a custom site maintainable after launch and makes handoff possible without technical hostage-taking.