About

Written under one rule: say what is true today, not what would be more impressive if it were true. A reader who later finds the gap between this page and reality is a reader we have lost permanently.

The page has three sections — who, what we have actually built, what we have not built yet — and one footer.

Section 1 — who

Leefan Reports is the publishing-operations practice of one named operator. At v1 launch, the practice is the operator plus contracted help on specific deliverables. It is not a multi-person agency wearing an agency name.

The operator’s working profile, as of 2026-05:

Has run a card-driven Claude Code operations loop on an internal publishing operation during the 2026-05 launch cycle. The operation itself is not linked from this page because it is not part of the clean-brand surface; the artifact this practice publishes here is the method — cards, regression DB shape, diagnostics, and rollback conventions.
Has shipped enough pages through that loop to expose real operational failures: duplicated templates, stale reports, broken measurement, and rollback decisions. The public claim is the method and the diagnostics, not the identity of the underlying operation.
Is the same person who reads inquiries sent to the Leefan Reports address. There is no inbox triage layer, no SDR, and no agency-style account manager.

What the operator is not claiming on this page:

Not a former Google / Meta / OpenAI staffer. If a credential of that kind existed, it would be listed here. It is not listed here because it does not exist.
Not the author of a widely-cited paper or a popular open-source project. Future external recognition will be cited inline as it accrues.
Not currently operating an agency with a roster of named clients. The paid engagement shapes on /services are offered going forward, not retroactively. We do not list logos we have not earned.

The operator’s day-to-day work is the work shown on this site: cards, regressions, diagnostics. The /about page is short on biography because the biography is the library.

Section 2 — what we have actually built

Artifacts a reader can verify in 10 minutes.

An open card library at /opslab/: five pages on what a Claude Code “card” is, how to run one, how to roll one back, and how a regression DB sits beside it.
Three side-by-side tool comparisons at /toollab/: each page states its weighting up front, declares an explicit counter-case where the loser is the right choice, and labels its benchmark numbers honestly (T01 / S01 / M01 currently render as numeric rerun pending or rubric-only here).
Six diagnostic and template pages at /diagkit/: a content-thinness self-check, a GA4 click_* firing checker, an affiliate compliance scan (held to v1.1 — coming after legal review), a Notion pSEO PM workspace, a card starter pack, and a decision-log template.
An evidence pack available by request at contact@leefan.co.jp: raw inputs to every benchmark on this site, the regression DB schema, the card schema, and a starter pack of five cards a reader can run today.

All of this is what we run on our own loop. The /about page does not have a separate “case study” carousel because the library is the case study — every page exists because it was load-bearing in a live operation.

Section 3 — what we have not built yet

A serious /about page tells the reader what is missing. The reader will find out anyway; better that they find out from us.

No paid-client case studies. Leefan Reports as a paid practice goes live at this launch. Until a client engagement completes, this page does not have a “results we delivered for X” section.
No revenue, ranking, or traffic guarantee. Anywhere on the site.
No team page. The practice is one named operator at v1. A team page will exist when there is a team; today, claiming one would be a lie.
No press / podcast / conference list. None has happened yet. When it does, it will be listed inline, dated, and linked to the actual recording or transcript.
No legal-counsel sign-off on the affiliate-compliance diagnostic (D04). That page is held from v1 launch and ships in v1.1 after a counsel pass.

What this section is not is false modesty. Underclaiming is just overclaiming in the other direction. The /opslab and /diagkit pages are operator-grade; we stand on them. The list above is the set of claims we cannot make truthfully today.

Section 4 — how to reach the operator

No contact form on /about by design. The two surfaces are:

contact@leefan.co.jp — engagement inquiries (read by the operator).
contact@leefan.co.jp — issues, corrections, and benchmark replication notes (read by the operator).

The contact address is rendered as plain text here. No form, tracker, calendar, or public repo URL is embedded. No scheduling link is dropped on /about; /services handles scheduling once an inquiry has been triaged.

Section 5 — what would change on this page over the next 12 months

When	Change to /about	Trigger
After first completed engagement	Add a single, opt-in client reference, with the client’s named results.	Engagement closes and client signs off on the wording.
After first independent benchmark rerun	Replace placeholder caveats on ToolLab pages with linked companion-repo runs; restate “verified” date on /about.	Rerun completes.
After legal-review pass on D04	Move D04 from “held” to “available”; remove the “v1.1” line.	Counsel signs off.
When a second operator joins	Add a small team section, named, with role boundaries.	A second person actually starts running the loop.
When a paid case study is shippable	Add an inline reference here and a full case study under `/case-studies/`.	Engagement completes and client opts in.

If a reader returns to this page in six months and the page has not changed, the practice has not changed.