A self-directed, AI-assisted work sample for DTC marketers and agency teams reviewing email before platform testing. This static-checking module is part of the Repeat Foundry Email Preflight repository, alongside a separate synthetic lifecycle decision lab.
The problem
Unfinished links, missing tracking fields, image-alt omissions and selected personalization risks can survive a copy review. A small offline preflight can make those issues easier to inspect before the account owner performs native testing.
The work
A standard-library Python command-line tool reads a local JSON manifest and HTML email as text. It returns stable rule codes, severity levels, locations and manual-review reminders. It does not render or execute the HTML, call a network service, use credentials or send email.
Worked example
The deliberately unfinished welcome-email fixture has 4 errors and 4 warnings. The reviewed fixture has 0 errors and 0 warnings within the implemented static checks. Both the before/after HTML and JSON reports are visible in the repository.
The examples are fictional. Those counts describe the seeded fixtures only; they are not measured client outcomes or proof of deployment readiness.
Handoff boundary
A clean report is not send approval. A real sender footer, platform preview, consent and eligibility checks, live-link tests and accountable owner approval are still needed. The tool does not establish deliverability, legal compliance, rendering quality or live flow behavior.