Semantic Atomicity Review
Generated artifact. Do not edit manually.
Generated by:
scripts/build_compendia.pyRegenerate with:python3 scripts/build_compendia.py --compendium qa-ci --web-and-texSource of truth:docs/qa-ci/source/
Language / Sprache: English (EN) · Deutsch (DE) | PDF: English PDF · Deutsches PDF
Compendium: §1 Maturity & Trials · §2 Scopes & CQR Rules · §3 Graph Validation · §4 Semantic Atomicity · §5 Champion Guide · §6 Four-Level Model
4. Semantic Atomicity Review
Rule version semantic-atomicity-v1
This QA lane verifies whether technical content leaf goals (empty contains) are also semantically atomic (CQR-301 in 2.3). Tooling tracks coverage and fingerprint freshness only; the atomicity judgment itself is strictly semantic rather than algorithmic.
4.1. Domain Criterion, Scope Exclusions, and Review Statuses
Under semantic-atomicity-v1, a technical content leaf goal is semantically atomic if and only if it expresses one cohesive assessable competence rather than a bundle of independent routines (conjunctions such as "calculate and interpret" act as review signals rather than automatic failures). Non-content leaves are intentionally excluded and validated in their dedicated lanes:
- terminal practice and assessment nodes (
Practice,Assessment, orexamData), - motivation and orientation anchors (
Motivation,Orientation), - memorization and SRS deck leaves (
memorizationorsrs-deck:*).
| Review status | Meaning in the JSONL ledger |
|---|---|
atomic |
Reviewed and accepted as semantically atomic (semanticAtomic: true). |
needs_developer_review |
Escalated for closer subject/developer review (semanticAtomic: null). |
non_atomic |
Too broad; scheduled for splitting (semanticAtomic: false). |
4.2. Active Configurations, Fingerprints, and Expansion Workflow
Configuration files and JSONL review ledgers reside under curricula/DE/Gymnasium/quality/semantic-atomicity/, for example the pilot lane canonical-math-j8-linear-functions-pilot, the foundational lane canonical-chemistry-sek1-foundations (with 26 accepted atomic leaves), and the full-subject lanes canonical-physics-full and canonical-math-full. In the local Workbench, route /semantic-atomicity-review provides an interactive status editor while CLI scripts run the checks:
cd app
npm run quality:semantic-atomicity
npm run quality:semantic-atomicity:check
quality:semantic-atomicity:check fails on process violations (missing in-scope record, stale fingerprint after semantic edits, malformed or duplicate rows), whereas known needs_developer_review and non_atomic items remain visible in the CQR-301 (M5) queue. Every record stores a semantic fingerprint over title/titleEn, description/descriptionEn, shortKey, phase/area/topic code, and nodeKind (--write-fingerprints may be run solely as final bookkeeping after an actual semantic review).
Six steps to onboard a new subtree: (1) add a .config.json under curricula/DE/Gymnasium/quality/semantic-atomicity/, (2) reference the canonical landscape and subtree roots, (3) initialize .review.jsonl with one record per content leaf, (4) run npm run quality:semantic-atomicity -- --config=<path>, (5) write --write-fingerprints once semantic decisions are recorded, and (6) wire the config into CI only after the initial batch is complete.