Open operating decisions

The following remain deliberately unresolved until real content makes the trade-offs concrete:

• detailed templates for Briefing and Topic Review, and any additional Paper Review sections beyond its adopted Main text, optional Supplementary information, and References structure; the settled pieces are the date-based Briefing title (Month D, YYYY, replacing the earlier Week N, YYYY form as of 2026-08-08), per-item subject-code citation-key conventions, and the Paper Review rule that Main text follows the reviewed paper’s section order while Supplementary information selectively holds useful source-supplement detail,
• whether reviews/paper/ or reviews/topic/ should split by subject once a real corpus makes those index pages useful (distinct from the per-topic subfolder pattern resolved 2026-08-12 below, which splits one Topic Review’s own content rather than the index),
• the broader frontmatter schema within Concepts beyond the resolved canonical subject-folder convention,
• whether the working mode and relation descriptors should become public metadata or taxonomy (field itself is resolved — see Recently resolved below),
• dedicated Briefing/Review filter routes (e.g. a real /briefings/quantum page) until a real approved corpus demonstrates the need — the lightweight client-side subject/date filter added 2026-08-10 is not a route and doesn’t resolve this,
• correction-log presentation,
• the detailed Study-to-Concepts promotion procedure — trigger, frontmatter carryover, and handling later edits to the source note (see Study-to-Concepts promotion for what is already fixed),
• Phase 1b export and editorial automation,
• renaming sungmin-park-dev/project-emergence to reduce its remaining naming clash with this project — deferred until that project’s own scope settles; don’t rename it as a side effect of work here, and
• the candidate-scoring weights (equal initial weight to curiosity, importance, and transferability) — revisit with user feedback after the first three published Briefings or Reviews,
• whether/when the /reviews/paper and /reviews/topic indexes need richer presentation; their current role is a simple page-tree mirror.

Recently resolved: Briefings, Reviews, and Concepts are separate reader-facing collections. Reviews explicitly distinguishes Paper Reviews of one contribution from Topic Reviews of a multi-source research topic, theory, or research program. Knowledge is not a separate collection, and the provisional Exposition role is retired. Project and collection explanations live under /about/**; actual knowledge content lives under /briefings/**, /reviews/**, and /concepts/**; editorial operations and technical maintenance have separate top-level branches. Every content folder has an authored index.md, and old routes remain compatibility aliases.

The emergence-hero/custom.scss split ownership debt was also removed on 2026-08-06. Component-owned .emergence-* DOM styles now live with the component, while route shell and Quartz-global styling live in custom.scss; see Local components. The same pass adopted a Quartz/Preact/SCSS-specific UI review profile rather than copying UI Skills’ Tailwind/React assumptions wholesale.

The automated daily research routine was also built on 2026-08-06 as scheduled task emergence-daily-candidates (/Users/david/.claude/scheduled-tasks/emergence-daily-candidates/SKILL.md, daily 8am): it finds ~10 candidates across the site’s full subject scope, proposes them and waits for the user to pick which to draft, then drafts selected item(s) as unlisted: true and leaves changes uncommitted for human review — never auto-commits, never auto-publishes. It replaces the old daily-study-digest task (deleted; its SKILL.md left on disk for reference), which wrote to the private Study vault instead of this repo.

Multi-item Briefing titling moved from Week N, YYYY to the plain publication date, Month D, YYYY, on 2026-08-08, once the daily-candidates task’s first real run produced a same-week supplemental roundup and Week 32, 2026 stopped uniquely naming a page. 2026-w32.md was renamed 2026-08-06.md (title August 6, 2026) with the old slug preserved as an alias. The earlier 2026-w31.md candidate page was removed on 2026-08-09 and is not retained as a Briefing, Review, or public format example. About Briefings instead uses the 2017 Transformer paper “Attention Is All You Need” to explain the format boundary. A Briefing may contain one or several selected items; a Paper Review is distinguished by substantially deeper source-centered examination, not merely by containing one paper. See Briefing and Review authoring for the current convention.

The Paper Review base structure was resolved on 2026-08-11 from the first Carleo–Troyer draft. Main text contains an editorially labeled Introduction and the paper’s own sections in source order as numbered subsections. Optional Supplementary information follows only when source-supplement derivations, implementation details, solver or cost analysis, additional tables, or convergence checks materially help the Review; central evidence and important limits remain in Main text. References closes the document. Additional editorial sections remain open until completed Reviews demonstrate a stable need.

The field taxonomy was resolved 2026-08-10, replacing the earlier open/provisional list. Seven reserved values: Quantum Information & Computing, Quantum Sensing & Metrology, Many-body/Materials, ML/AI, Numerical Methods & Optimization, Electronic Devices & Engineering, Mathematics — full definitions, the goal-not-toolkit assignment rule, and the straddle/venue tests used to decide the boundaries are in DESIGN.md. This went through two rounds of independent model review (Opus, then ChatGPT Pro) specifically because it also constrains the future concepts/ folder tree and the private Study vault, not just Briefing/Review tagging. Quantum Sensing & Metrology held its first item in the 2026-08-13 Briefing (a ## Quantum Sensing & Metrology section covering an indefinite-evolution multiparameter metrology paper) — one of its two original population triggers (an edition needing its own ## section for it), while the other (~3 sensing/metrology items existing) has not yet fired. Quantum Information & Computing itself is expected to eventually split further (most likely sensing first, per the same trigger) once real volume exists; this is an accepted future move, not a sign the taxonomy is wrong today. A same-day lightweight client-side subject+date filter shipped on /briefings alongside this decision — see Site maintenance / emergence-hero’s collection-index.tsx.

The Concepts folder tree was aligned to the same subject set on 2026-08-10. Six subject routes contained real pages at that time; quantum-sensing-metrology remained reserved without an empty folder until 2026-08-13, when its first Concept (quantum-metrology, linked from that day’s Briefing) and an authored subject index gave it a reader-facing purpose — all seven reserved subject routes are now active. Each Concept has one canonical subject home, while cross-field relationships stay in its metadata and links. This resolves the per-entry path convention but does not freeze the broader Concepts frontmatter schema.

The Concept quality contract and repository template were resolved on 2026-08-10 after the existing 43 non-index pages were reclassified as a rewrite queue rather than accepted examples. A Concept must define and bound one reusable idea, expose the structure needed to reason with it, interpret its use, state limits and confusions, connect genuine neighboring knowledge, and match substantive claims to sources. Exact headings and length remain subject-dependent. Quantum state tomography is the first rewritten pilot; the remaining pages have provisional rewrite or routing dispositions in Concept rewrite register. This decision does not approve any Concept for publication or settle the broader metadata schema.

The Topic Review per-topic subfolder pattern was resolved 2026-08-12. A Topic Review defaults to a single file, reviews/topic/{slug}.md. It may graduate to a subfolder, reviews/topic/{slug}/, once its reconstruction needs standalone pages for formal apparatus specific to that research program’s own literature — apparatus reusable enough across contexts to matter beyond the one program still belongs in the general Concepts collection instead. The reconstruction itself moves to {slug}/index.md (unchanged route, still the only entry the /reviews/topic hub lists); each apparatus page is a sibling file reachable only by link from that index, carries no collection/format/date/field, and keeps unlisted: true until reviewed and approved like any other page. entanglement-bootstrap/ is the first Topic Review adopting this — see DESIGN.md’s Information architecture and Research content storage and routes. This does not resolve the separate, still-open question of splitting the reviews/topic/ or reviews/paper/ index by subject.

See Operations and publishing for the full operating model.