Cloudflare infrastructure
Around twelve sessions repairing live sites, diagnosing build failures, and configuring HTTPS/HSTS with Universal SSL, plus custom domains onboarded through TransIP DNS.
Blog / AI organisation
A good day with Claude
What 685 messages, 43 sessions and 407 hours of directing Claude Code actually looked like — and how a local usage report turns that history into an improvement loop.
A report that reads your own history back to you
Every fortnight a report lands quietly in a local folder: no dashboard, no login, no vendor watching over a shoulder. Just a summary of what actually happened across two weeks of working with Claude Code. This one covered 685 messages across 43 sessions, from 23 June to 6 July 2026.
It is worth reading not because the numbers are impressive on their own, but because of what they describe: a working pattern that has quietly shifted from asking a tool for snippets to directing an operator through full delivery.
The fortnight in numbers
- 685 messages across 43 sessions
- ~407 hours of session time
- 201 commits shipped
- 3,000+ Bash calls executed
Five workstreams, two weeks
The work was not a demo. It was real, parallel delivery.
Platform feature development
Passkey auth, clipboard tracking, a comment-overlay plugin system, and a full user model backed by D1 and identity — including a Chrome extension debugged through several deploy-test-fix cycles.
Documentation site build-out
Gap analysis, directory expansion, monitoring and alerts, a UI shell, personalised pages, and a full product rename across every repo in scope.
RBAC and tenancy architecture
A multi-phase admin build, a role-based access rollout with production migrations, and a per-customer tenancy model spanning apps, pricing and provider settings.
Documentation and repo restructuring
Worktree-based workflows, sub-projects extracted into their own repos and submodules, branches reconciled onto main, and codebases restructured into monorepos.
What made it a good day, not just a busy one
End-to-end ship-and-verify, not ship-and-hope. On a role-based access control rollout, the work did not stop at a green build. The production migration was applied, then the fix was proven by recovering a one-time password and confirming a live sign-in.
Multi-day epics that actually land. One platform upgrade ran as a four-day plan: features implemented, deployed to production, 500 businesses seeded as real data, and day five queued up before the two weeks were out.
Discipline that survives a real multi-repo setup. Worktrees, submodules, PR-based merges and structured task tracking held up across 201 commits without turning into chaos.
“Verification is part of ‘done’, not an afterthought.”
Closing note
Direct clearly, verify honestly, and let the tool run.
None of this reads as “the AI wrote some code”. It reads as a technical lead directing a very fast, very literal engineer who needs the plan stated clearly and then gets out of the way.