v2 phase 1: TiCS scaffold via manuscript-formatting

[neuromechanist]

Summary

v2 Phase 1: scaffold a TiCS Forum Review by actually invoking the manuscript:manuscript-formatting skill via the Skill tool. The skill loaded its full instructions and I applied them.

What the skill provided

• Submission checklist (title page, abstract, figures at 300+ DPI, references consistent, conflict of interest, data availability, CRediT, suggested reviewers, etc.).
• BibTeX management guidance (Format Conversion section + plain/unsrt/ieeetr/apalike styles).
• LaTeX/Word/Markdown conversion via pandoc.
• Cover-letter template.
• references/journal-requirements.md table (IEEE, Nature, Science, PNAS, PLoS ONE, Elsevier, Frontiers, NeuroImage).

What the skill did NOT provide

• TiCS-specific Forum Review scaffolding (Highlights, Trends Box, Outstanding Questions Box, Glossary). These are TiCS conventions not encoded in the skill or its journal-requirements reference.
• An automated per-section file generator. The skill is a methodology guide.

TiCS-specific decisions applied on top of the skill

(Sourced from public TiCS author guidelines, documented in STRUCTURE.md.)

• Forum Review main text ~3500-4000 words.
• 3-5 Highlights, <=85 chars each (Elsevier house style row in the skill's journal-requirements doc).
• Trends Box (~200-250 words).
• Outstanding Questions Box (5-7 questions).
• Glossary (10-15 defined terms).
• Abstract ~80-120 words, no citations.
• Cell Press numbered references (Vancouver-like).
• No IMRAD (TiCS Forum Reviews are thematic).

Files created

manuscript/narrative-review/
  STRUCTURE.md                # TiCS compliance map + skill checklist tracking
  frontmatter.yaml            # title, authors, affiliations, word budget
  highlights.md               # 5 highlight placeholders, <=85 chars each
  abstract.md                 # ~120-word abstract placeholder
  sections/
    01_introduction.md        # ~450 words budget
    02_scaffold.md            # ~600 words budget
    03_psychophysics.md       # ~500 words budget
    04_action.md              # ~500 words budget
    05_language.md            # ~600 words budget (5a + 5b split per v1 I5)
    06_emotion.md             # ~500 words budget
    07_synthesis.md           # ~450 words budget
  boxes/
    box1_anchor.md            # HBN-EEG R3 anchor cohort (F4 reframe)
    trends.md                 # 5-7 recent developments
    outstanding-questions.md  # 7 forward-looking questions
  glossary.md                 # 15 defined terms placeholders
  figures.md                  # figure-legend placeholders + Phase 3 composer notes
  references.md               # Phase 5 build notes; F2 carry-forward documented
  refs.bib                    # 94 BibTeX entries (deduplicated; F3 stray-cleaned)

F1-F5 carry-forwards from v1 self-review

Documented in STRUCTURE.md and individual section file headers:

• F1 (cardinality 12 -> 7): Section 5a placeholder names the verbatim phrasing.
• F2 (Schubring vs Codispoti DOI): refs.bib header + references.md note.
• F3 (stray refs removed): applied to refs.bib (94 entries, was 97).
• F4 (anchor case reframe): Section 7.2 + Box 1 placeholders restructured.
• F5 (falsifiability rejection region): Section 7.3 placeholder.
• I1-I5: per-section file headers.

Closes #9. Part of the v2 narrative-review redo.

Test plan

• Skill loaded and applied via the Skill tool (not paraphrased)
• Skill outputs vs gaps documented honestly in the PR body
• 17 scaffold files created at expected paths
• refs.bib has 94 entries after dedup + F3 stray removal
• Pre-commit (markdownlint, typos) passes
• Phase 2 (manuscript-writing) will fill placeholders per section
• Phase 3 (figures:scientific-figure composer) will build the 4 figures
• Phase 4 (paper-review + humanizer) will review
• Phase 5 (final assembly) will compile and open the final PR