b21 launch-6 — Supplementary Information template + stamp across series#

Mode: EDEN
Effort: Max
Run order: 6 of 7 in the b21 launch-sprint sub-prompt set. Post-launch.
Status of upstream: SI template scoped in b20 review llog §7-3 + §5.6.
Drafter: Claude Opus 4.7 Max (dv_ClaOp47Max) at LLoL’s request on 2026m05d22.

1. Context#

Each Matheo paper (b11–b21) has a Supporting Information / Appendix section that handles AI co-authorship rationale, Methods AI use, Acknowledgements, project metadata, and cross-reference manifest. The structure varies across papers; the b19 SGIR paper and b21 AI framework paper have the most developed versions.

The goal of this prompt: extract a canonical SI template from those two reference papers, lock it in an AHA file, then stamp the template across b11–b18a so every paper has the same SI structure (with paper-specific content within sections).

This prompt has two passes in one session:

  • Pass 1: Define the template (read b19 + b21 SI; extract skeleton; write AHA file; lock).

  • Pass 2: Stamp the template across b11–b18a. (Long; see timing §6.)

2. Pass 1 — Template definition#

2.1 Read source patterns#

Read the SI / Appendix sections of:

  • b19 SGIR (source/matheology/hell/mm/b/19/...) — the locked-pattern paper.

  • b21 AI framework (source/matheology/hell/mm/b/21/...) — the framework paper.

For each, identify:

  • Section headings in order.

  • Approximate length per section.

  • What content lives where (AI rationale, Methods, Acknowledgements, project metadata, cross-reference manifest).

  • Any sections unique to one paper that should NOT be in the generalised template.

2.2 Extract the canonical skeleton#

Define a minimal-but-complete SI section skeleton. Suggested starting structure (tune to match what b19 + b21 actually have):

Supporting Information
========================

SI.1 AI co-authorship rationale
---------------------------------
(Per AHA standard SI paragraph form --- ~170 words.
Cross-ref to title-page fn 7.)

SI.2 Methods --- AI use
--------------------------
(Per AHA standard Methods form --- ~50 words.)

SI.3 Acknowledgements
-----------------------
(Paper-specific; closed form.)

SI.4 Project metadata
-----------------------
(Stable URL, repo, build chain.)

SI.5 Cross-reference manifest
--------------------------------
(Related papers in the Matheo series; data; figures.)

Adjust per what b19 + b21 actually show.

2.3 Write the AHA file#

Write the template to AHA/study-supplementary-information-template.md.

Include:

  • The canonical skeleton (with locked section ordering).

  • Per-section length guidance.

  • Per-section content rules (what goes here; what does NOT).

  • A rendering pattern (RST snippet showing the headings + first paragraph of each section as a template).

  • A “Checklist for stamping the SI template onto a new paper” modeled on the AHA’s footnote-template checklist.

2.4 Lock and confirm#

Present the AHA template draft to LLoL for confirmation before Pass 2 begins. If LLoL approves: lock the template and proceed. If LLoL requests changes: revise; re-present; loop until approved. Do NOT proceed to Pass 2 with an unlocked template.

3. Pass 2 — Stamp across b11–b18a#

3.1 Target papers#

Same canonical-file identification as launch-5: b11, b12, b13, b14, b15, b16, b17, b18a.

3.2 For each target paper#

  1. Locate the paper’s current SI / Appendix section.

  2. Compare against the canonical skeleton from Pass 1.

  3. For each canonical section missing: add it (with paper-specific content where needed; defer to LLoL for any non-mechanical content additions).

  4. For each existing section that doesn’t match the canonical: flag in llog; do NOT rewrite without LLoL confirmation.

  5. Update the section ordering to match the canonical.

3.3 Cross-paper consistency#

After all papers in scope are stamped:

  • Read the SI sections of all stamped papers in sequence.

  • Verify section ordering is consistent.

  • Verify the AI co-authorship rationale text is the AHA-canonical ~170-word paragraph (paper-agnostic).

  • Verify the cross-reference manifest format is consistent (links rendered the same way).

3.4 Build test#

make dev. 0 new warnings target.

4. Out of scope#

  • b19, b20, b21 (already match the canonical, by definition — they ARE the canonical).

  • b18-esc Beginner / Producer / Expert (still in study-llog).

  • Any formal content in main text (axioms, theorems, body) — SI section only.

  • Generating paper-specific Acknowledgements text (paper-author’s job; defer to LLoL for any new content).

5. Deliverables#

Produce two artefacts:

5.1 Pass 1 deliverable#

A new file at AHA/study-supplementary-information-template.md with the canonical SI skeleton, rendering pattern, and stamp checklist.

5.2 Pass 2 deliverable#

A new LLog file at source/matheology/hell/ll/infra/b/NN/bNN-launch-6-execution_<date>.rst containing:

  • Pass 1 template summary (link to the AHA file; brief on what was locked).

  • Per paper: before/after diff of SI section structure (highlighting added sections, flagged inconsistencies, deferred content additions).

  • Build test result.

  • Cross-paper consistency report.

  • Concluding summary.

Also update the infra index.

6. Timing & BABL-resistance#

Target: 90–120 minutes for Pass 1 (template) + 60 min per paper for Pass 2 stamp = 6–10 hours total.

This is the longest of the 7 launch-sprint prompts. If the session is running long, split: complete Pass 1 (template + AHA file) in one session; stamp Pass 2 across papers in a follow-up session (or one paper per session). Either way, Pass 1 must finish cleanly before Pass 2 begins.

The BABL trap: over-Reaching into content rewrites. Pass 2 is structural alignment, not content polishing. If a paper’s SI section looks dated or thin, flag in llog for LLoL to schedule a content-pass — don’t fix inline.