paperlab-scientific-traceability-audit

PaperLab Scientific Traceability Audit

When to Use

USE WHEN: a textbook chapter contains numbers, equations, figures, tables, or engineering recommendations that depend on calculations, simulations, standards, or source documents. This skill turns technical content into auditable content.

Pair with:

• neqsim-in-writing for NeqSim claim markers,
• neqsim-notebook-patterns for executable notebooks,
• neqsim-regression-baselines for stable examples,
• neqsim-standard-requirement-extraction for clause-level standards evidence.

Traceability Objects

Track six object types:

Object	Required Evidence
Numeric claim	notebook cell, test, baseline, source table, or citation
Figure	source file, notebook, generated script, or approved source document
Equation	symbol definitions, units, source citation, and implementation link if used by NeqSim
Table	data source, units, and update path
Standard statement	standard code, revision, clause or paraphrased requirement
Recommendation	preceding evidence that supports the action

Marker Rules

Use existing neqsim-in-writing markers for high-value NeqSim-backed content:

<!-- @neqsim:claim
  test: src/test/java/neqsim/book/<book_slug>/<TestName>.java
  baseline: src/test/resources/baselines/book/<book_slug>/<baseline>.json
-->

<!-- @neqsim:figure source=notebooks/<notebook>.ipynb cell=<cell-id-or-index> -->

<!-- @neqsim:eq method=neqsim.package.ClassName#methodName -->

Do not mark every sentence. Prioritize values and equations that students will reuse, cite, or compare in exercises.

Audit Workflow

1. Run book-skill-stack-plan and inspect traceability dimensions.
2. Run book-evidence-check and book-figure-dossier.
3. For each core chapter, identify the top 3 to 5 numerical or equation-based claims that deserve markers.
4. Confirm every cited notebook exists and has executed output or generated figures.
5. Confirm units in prose, tables, and figures match nomenclature.yaml.
6. For standards-heavy chapters, add clause-level or paraphrased evidence rather than broad standard names.