# Editorial practice packet

Six new complete tasks and one deeper existing survey guide. Seven bilingual example/record pairs give 28 Markdown files. The six new guides contain 18 constructed cases. The survey retains its three questions and adds six explicit routing paths and three neutral post-answer probes; the separate registered reader study is unchanged.

Full original examples include three voice/tone drafts, a complete request/file explainer, a bounded opinion with an objection, two neutral usability tasks, an eight-unit descriptive codebook and a complete fictional editorial-policy specimen. No real person, observer, independent reviewer, test session, policy adoption, native post or payment is represented.

Run node check-examples.mjs with an existing Node installation. It checks six declared request events/two file IDs, overlapping codes across eight units, retained contrary/unknown states, and the declared survey routing. These checks do not interpret arbitrary prose, certify methods or establish rights, consent, comprehension, accessibility, independence or business outcomes.

render-coding-map.mjs regenerates two localized SVG files for one original diagram design from practice-fixtures.json using Node built-ins and no network. Labels and a complete HTML/Markdown table preserve the same meaning. Five assignments across four coded units are not five people; the fixture has no independent coders or measured reliability.

Primary bodies were read on 1 November 2026 in their declared scope. Mailchimp rules describe its own publication; Purdue guidance concerns academic argument, not a universal post formula. GOV.UK analysis is adapted transparently, not a universal thematic-analysis method. AP attribution/correction practices do not grant affiliation or certify our policy. AAPOR pretesting is not performed by an arithmetic check. Purdue was read through its actual official web body after direct retrieval error; SPJ HTML/PDF403 are excluded despite cached search coverage. No private intranet or underlying study data was accessed.

Keep completed records minimal and private. Unknowns remain unknown. The identity manifest excludes itself and ZIP against circular hashes; the archive contains every source file plus that manifest. No new research piece, aid category, independent review or reader/customer credit is claimed.