feat: Link-Taxonomie abgeschlossen - Lint hart ab kb_version 4.0.0, outbound: an Type-Spec gebunden, part-of/composition als Inversenpaar (4.1.0, #40)

Files changed: - CHANGES.md - VERSION - instructions/link-taxonomy.md - instructions/migrations/4.0.0-link-taxonomy.md - kb/comparisons/amd-pstate vs acpi-cpufreq.md - kb/concepts/Episodic Memory.md - kb/concepts/Memory Lifecycle.md - kb/concepts/Mesh Sync.md - kb/concepts/Procedural Memory.md - kb/concepts/Reciprocal Rank Fusion.md - kb/concepts/Semantic Memory.md - kb/concepts/Shared vs Private.md - kb/concepts/Split Threshold.md - kb/concepts/Stub Threshold.md - kb/concepts/Supersession.md - kb/concepts/Typed Relationships.md - kb/concepts/Vector Search.md - kb/concepts/Work Coordination.md - kb/concepts/Working Memory.md - kb/entities/technologies/Wine-Staging.md - kb/entities/tools/pascalandy schema.md - kb/index.md - kb/log.md - kb/sources/COLLECTION.md - tools/CONTRACT.md - tools/chemenu/commands/lint.py - tools/chemenu/kb_collections.py - tools/chemenu/lint_core.py - tools/chemenu/tests/test_conventions.py - tools/chemenu/tests/test_lint.py - tools/chemenu/tests/test_type_resolver.py - types/comparison.md - types/comparison.schema.yaml
2026-09-03 06:19:56 +02:00
parent 23e34a940c
commit cfe925a76c
33 changed files with 379 additions and 66 deletions
@@ -20,6 +20,68 @@ their date-only headings.

 ---

+## 4.1.0 - 2026-09-03 - Link-Taxonomie: Lint-Findings hart ab kb_version 4.0.0, outbound: an das Type-Spec gebunden, part-of/composition als Inversenpaar
+
+**Author:** Torben Nehmer
+
+Der Rest von Issue #40, nachdem die Korpus-Migration durch ist: die beiden aufgeschobenen
+Lint-Findings werden hart, und die drei Befunde aus dem Abschlusskommentar des Migrationslaufs
+werden aufgelöst.
+
+**`unlabelled_edges` und `unauthorised_labels` sind harte Fehler — aber an `kb_version`
+gebunden, nicht an ein Datum.** Der Weg, den `legacy_citation_markers` genommen hat, war ein
+Umlegen in einer späteren Version: eine Instanz, die die Zitat-Migration noch schuldete, lebte
+danach mit rotem Lint. Das Ledger kann die Frage inzwischen beantworten, also tut es das.
+Unterhalb `kb_version` 4.0.0 bleiben beide beratend — genau das Fenster, in dem
+`instructions/migrations/4.0.0-link-taxonomy.md` der Instanz sagt, sie solle den halb
+konvertierten Korpus Einheit für Einheit publizieren; ein Check, der dabei fehlschlägt, würde
+den Korpus verweigern, dessen Fortschritt er misst. Ab 4.0.0 ist eine kahle Titelangabe in
+`related:` keine Seite mehr, die auf ihre Umstellung wartet, sondern eine Kante, deren Autor
+nicht gesagt hat, was sie behauptet. `hard_error_keys()` liefert die jeweils geltende Menge,
+`HARD_ERROR_KEYS` bleibt die vollständige.
+
+**`outbound:` ist an das Type-Spec gebunden.** `kb/sources/` und `kb/comparisons/`
+autorisierten Label, die dort strukturell nicht schreibbar waren: keiner der beiden Type-Specs
+führte ein `related:`. Folgenlos war das nicht — die einzige Comparison-Seite des Korpus trug
+`- **compares-with:** [[amd-pstate]]` als *handgeschriebene Prosa*, ohne Marker-Region, ohne
+Frontmatter, für `lint` unsichtbar. Also ein Identifier zurück im Fließtext, gut vier Stunden
+nachdem 4.0.0 genau das beendet hatte. Eine leere Autorisierung liest sich als Lizenz.
+
+Aufgelöst nach dem, was die beiden Contracts jeweils selbst sagen: `comparison` bekommt ein
+`related:` (die `compares-with`-Kante gegen jedes Subjekt ist die eine Aussage, für die die
+Seite existiert), `kb/sources/` verliert seinen `outbound:`-Block ersatzlos (dessen Contract
+sagt ausdrücklich, seine Verknüpfungen seien der mechanische Provenance-Pfad und keine
+Autorenkanten). Neu prüft `docs verify` die Kombination: ein `outbound:`-Block auf einer
+Collection, in die kein Typ mit `related:` schreibt, ist ein Befund und nennt beide Richtungen
+der Reparatur.
+
+**`composition` / `part-of` ist das dritte Inversenpaar**, neben `depends-on` / `required-by`
+und `runs-on` / `hosts`. Aus der Messung, nicht vom Schreibtisch: der u3-Lauf hatte entschieden,
+die Gegenseite eines `composition` bekomme `see-also`, weil `part-of` ein Spiegel wäre. Ist es
+nicht — der Satz des Elternteils zählt seine Teile auf, der des Kindes benennt das Ganze, zu
+dem es gehört, und ein Leser, der auf dem Kind landet, braucht den zweiten. Übrig blieben 16
+`see-also`-Kanten für eine Beziehung, für die der Katalog ein Wort hat; sie sind auf `part-of`
+umgestellt. Ein Inversenpaar macht die Gegenkante weiterhin **nicht** zur Pflicht — Richtung
+wird verfasst, nicht gespiegelt —, es legt nur fest, welches Label sie trägt, wenn jemand sie
+schreibt.
+
+**Stack- und Korpusänderung laufen hier in einem Zug**, entgegen der sonstigen Trennung. Der
+neue `docs verify`-Check würde eine bestehende 4.0.x-Instanz beim bloßen Kopieren der neuen
+Maschinerie fehlschlagen lassen, weil deren `kb/sources/COLLECTION.md` den `outbound:`-Block
+noch trägt — nach [instructions/dev/version-parts.md](instructions/dev/version-parts.md)
+Schritt 1 ein Grenzübertritt. Statt dafür eine `5.0.0` zu lösen, ist die Ursache mitbeseitigt:
+die Collection-Contracts dieser Instanz sind angepasst, und `dist export` leitet die
+`COLLECTION.md.template` daraus ab, also liefert jede neue Distribution die korrigierte Form
+aus. Für eine bereits bestehende 4.0.x-Instanz bleibt eine Handbewegung übrig, und sie wird
+hier benannt statt versteckt: die zwei `outbound:`-Zeilen aus `kb/sources/COLLECTION.md`
+löschen. Das neue `related:` im `comparison`-Type-Spec erreicht sie ohnehin nicht — die vier
+Page-Type-Specs gehören seit 4.0.0 der Instanz und werden nur als `.template` ausgeliefert.
+
+Offen aus #40 bleibt nichts mehr; Befund 2 des Migrationslaufs (dem Katalog fehlt ein Register
+für Urheberschaft) ist als eigenes Issue erfasst.
+
+---
+
 ## 4.0.1 - 2026-09-02 - Issue-Board: vier Pflicht-Label-Familien und Body-als-Wahrheit

 **Author:** Torben Nehmer
@@ -1 +1 @@
-4.0.1
+4.1.0
@@ -40,9 +40,18 @@ edge merely to mirror the first one.** The inbound view is rendered from the gra
 `index rebuild` and `search`, so a reader landing on the target sees what points at it whether
 or not anyone wrote a second edge.

-That is why most labels below have no inverse. Only two pairs do, because in each the reverse
+That is why most labels below have no inverse. Only three pairs do, because in each the reverse
 direction is a genuine primary statement someone would write on its own: `depends-on` /
-`required-by` and `runs-on` / `hosts`.
+`required-by`, `runs-on` / `hosts`, and `composition` / `part-of`.
+
+The third was added after the 4.0.0 migration, from measurement rather than from the desk. A
+parent-child structure - a tier list and its tiers, a spectrum and its levels - produces the
+question on nearly every page: the parent writes `composition`, and the child then reaches for
+either `part-of` or `see-also`. The migration run answered `see-also`, on the reading that
+`part-of` would be a mirror, and left sixteen edges saying "these two are related" about a
+relationship the catalogue already had a word for. It is not a mirror: the parent's sentence
+lists its parts, the child's names the whole it belongs to, and a reader landing on the child
+needs the second one.

 ## When to run

@@ -129,6 +138,13 @@ Inference and comparison between ideas.
 | `composition` | is composed of the target |
 | `part-of` | is a component of the target |

+`composition` / `part-of` is the third **inverse pair**, alongside `depends-on` / `required-by`
+and `runs-on` / `hosts` in the operational register. Being a pair does not make the second edge
+obligatory - direction is still authored - it settles *which label* the second edge takes when
+someone does write it. The child of a `composition` writes `part-of`, not `see-also`: what it
+is a component of is a primary statement about the child, and `see-also` says strictly less
+about the same fact.
+
 `grounds` / `rests-on` is a genuine pair and both directions are primary statements; they are
 listed separately rather than as inverses because either page may legitimately carry only its
 own side.
@@ -36,8 +36,9 @@ you know the run is finished.
 **Nothing breaks while it is outstanding.** Unlabelled edges and undelimited regions are read,
 not rejected: `links.py` treats a bare title as an edge whose label is not declared yet, and
 `provenance.split_cite_block` falls back to the pre-marker layout. That is deliberate - a corpus
-has to stay readable while it is being converted - and it is why the two lint findings are
-advisory until step 6 promotes them.
+has to stay readable while it is being converted - and it is why the two lint findings stay
+advisory for as long as `kb_version` is below 4.0.0, which is exactly as long as this document
+is outstanding.

 ## Steps

@@ -102,7 +103,7 @@ advisory until step 6 promotes them.
   is otherwise silent: the region becomes ordinary prose and the next write appends a second
   one beside it.

-6. **Record it, then tighten the checks:**
+6. **Record it. The checks tighten themselves:**

   ```bash
   tools/wikitool lint          # unlabelled_edges and unauthorised_labels must be 0
@@ -110,8 +111,14 @@ advisory until step 6 promotes them.
   ```

   Only once `lint` reports zero of both is the run finished. The two findings are advisory
-   during the window and become hard errors afterwards - the same path
-   `legacy_citation_markers` took after the citation migration.
+   while `kb_version` is below 4.0.0 and hard from the moment `migrate done` records it -
+   nothing to flip by hand, and no window in which a half-converted corpus is refused by the
+   check that is measuring its progress.
+
+   Do not record the migration to silence the findings. The promotion is what makes the run
+   stick: after it, a bare title in `related:` is a hard error rather than a page still
+   waiting, so a corpus recorded early fails its next lint instead of quietly keeping the old
+   shape.

 ## How to tell a migrated page from an unmigrated one

@@ -3,8 +3,10 @@ type: types/comparison.md
 tags: [kernel, power-management, amd, cpu, driver]
 created: 2026-07-31
 entities: [amd-pstate, acpi-cpufreq]
-summary: "Vergleich zweier AMD-CPU-Power-Management-Treiber: CPPC-basiertes amd-pstate gegen\xFC\
-  ber ACPI-basiertem acpi-cpufreq."
+summary: "Vergleich zweier AMD-CPU-Power-Management-Treiber: CPPC-basiertes amd-pstate gegen\xFCber ACPI-basiertem acpi-cpufreq."
+related:
+  - compares-with: amd-pstate
+  - compares-with: acpi-cpufreq
 ---
 # Comparison: amd-pstate vs acpi-cpufreq

@@ -131,8 +133,9 @@ ls /sys/devices/system/cpu/cpu0/cpufreq/cppc_*

 **amd-pstate** stellt einen bedeutenden Fortschritt in der CPU-Energieverwaltung für AMD-Prozessoren dar und bietet fein-körnige Steuerung, bessere Effizienz und verbessertes Batterielebensdauer. **acpi-cpufreq** bleibt ein zuverlässiger Fallback und dient weiterhin älterer Hardware. Die Wahl zwischen ihnen hängt hauptsächlich von Hardware-Unterstützung und Kernel-Version ab, wobei amd-pstate die klare Präferenz für moderne AMD-Systeme ist.

+<!-- wikitool:links -->
 ## Beziehungen

 - **compares-with:** [[amd-pstate]]
 - **compares-with:** [[acpi-cpufreq]]
-
+<!-- /wikitool:links -->
@@ -5,7 +5,7 @@ tags: []
 created: 2026-08-02
 modified: 2026-08-29
 related:
-  - see-also: Consolidation Tiers
+  - part-of: Consolidation Tiers
 sources: []
 confidence: 0.50
 confidence_base: 0.50
@@ -41,5 +41,5 @@ TODO
 <!-- wikitool:links -->
 ## Beziehungen

- **see-also:** [[Consolidation Tiers]]
+- **part-of:** [[Consolidation Tiers]]
 <!-- /wikitool:links -->
@@ -5,7 +5,7 @@ tags: [memory, lifecycle, confidence, knowledge-management]
 created: 2026-07-26
 modified: 2026-08-29
 related:
-  - see-also: LLM Wiki Pattern
+  - part-of: LLM Wiki Pattern
  - see-also: Confidence Scoring
  - composition: Supersession
  - see-also: Consolidation Tiers
@@ -126,7 +126,7 @@ Basierend auf [[Agent Memory]]-Erfahrung:
 <!-- wikitool:links -->
 ## Beziehungen

- **see-also:** [[LLM Wiki Pattern]]
+- **part-of:** [[LLM Wiki Pattern]]
 - **see-also:** [[Confidence Scoring]]
 - **composition:** [[Supersession]]
 - **see-also:** [[Consolidation Tiers]]
@@ -6,7 +6,7 @@ created: 2026-08-02
 modified: 2026-08-29
 related:
  - exemplifies: Implementation Spectrum
-  - see-also: Multi-Agent Collaboration
+  - part-of: Multi-Agent Collaboration
  - evidenced-by: Source - LLM Wiki v2
 sources: []
 confidence: 0.50
@@ -46,6 +46,6 @@ TODO
 ## Beziehungen

 - **exemplifies:** [[Implementation Spectrum]]
- **see-also:** [[Multi-Agent Collaboration]]
+- **part-of:** [[Multi-Agent Collaboration]]
 - **evidenced-by:** [[Source - LLM Wiki v2]]
 <!-- /wikitool:links -->
@@ -5,7 +5,7 @@ tags: []
 created: 2026-08-02
 modified: 2026-08-29
 related:
-  - see-also: Consolidation Tiers
+  - part-of: Consolidation Tiers
 sources: []
 confidence: 0.50
 confidence_base: 0.50
@@ -43,5 +43,5 @@ TODO
 <!-- wikitool:links -->
 ## Beziehungen

- **see-also:** [[Consolidation Tiers]]
+- **part-of:** [[Consolidation Tiers]]
 <!-- /wikitool:links -->
@@ -5,7 +5,7 @@ tags: []
 created: 2026-08-02
 modified: 2026-08-29
 related:
-  - see-also: Hybrid Search
+  - part-of: Hybrid Search
  - exemplifies: LLM Wiki Pattern
  - evidenced-by: Source - LLM Wiki v2
 sources: []
@@ -45,7 +45,7 @@ TODO
 <!-- wikitool:links -->
 ## Beziehungen

- **see-also:** [[Hybrid Search]]
+- **part-of:** [[Hybrid Search]]
 - **exemplifies:** [[LLM Wiki Pattern]]
 - **evidenced-by:** [[Source - LLM Wiki v2]]
 <!-- /wikitool:links -->
@@ -5,7 +5,7 @@ tags: []
 created: 2026-08-02
 modified: 2026-08-29
 related:
-  - see-also: Consolidation Tiers
+  - part-of: Consolidation Tiers
 sources: []
 confidence: 0.50
 confidence_base: 0.50
@@ -43,5 +43,5 @@ TODO
 <!-- wikitool:links -->
 ## Beziehungen

- **see-also:** [[Consolidation Tiers]]
+- **part-of:** [[Consolidation Tiers]]
 <!-- /wikitool:links -->
@@ -5,7 +5,7 @@ tags: []
 created: 2026-08-02
 modified: 2026-08-29
 related:
-  - see-also: Multi-Agent Collaboration
+  - part-of: Multi-Agent Collaboration
 sources: []
 confidence: 0.50
 confidence_base: 0.50
@@ -43,5 +43,5 @@ TODO
 <!-- wikitool:links -->
 ## Beziehungen

- **see-also:** [[Multi-Agent Collaboration]]
+- **part-of:** [[Multi-Agent Collaboration]]
 <!-- /wikitool:links -->
@@ -5,7 +5,7 @@ tags: [split, threshold, lines, pages]
 created: 2026-08-03
 modified: 2026-08-29
 related:
-  - see-also: Content Quality Control
+  - part-of: Content Quality Control
  - see-also: Stub Threshold
  - see-also: Index Scaling
 sources: [Source - LLM Improvements Sonnet Analysis]
@@ -69,7 +69,7 @@ Split Threshold definiert die maximale Größe, die eine Wiki-Seite erreichen so
 <!-- wikitool:links -->
 ## Beziehungen

- **see-also:** [[Content Quality Control]]
+- **part-of:** [[Content Quality Control]]
 - **see-also:** [[Stub Threshold]]
 - **see-also:** [[Index Scaling]]
 <!-- /wikitool:links -->
@@ -5,7 +5,7 @@ tags: [stub, minimum, quality, lines]
 created: 2026-08-03
 modified: 2026-08-29
 related:
-  - see-also: Content Quality Control
+  - part-of: Content Quality Control
  - see-also: Split Threshold
  - see-also: Semantic Lint Automation
 sources: [Source - LLM Improvements Sonnet Analysis, Source - LLM Wiki v2]
@@ -80,7 +80,7 @@ It provides comprehensive information about the topic. It clearly exceeds the st
 <!-- wikitool:links -->
 ## Beziehungen

- **see-also:** [[Content Quality Control]]
+- **part-of:** [[Content Quality Control]]
 - **see-also:** [[Split Threshold]]
 - **see-also:** [[Semantic Lint Automation]]
 <!-- /wikitool:links -->
@@ -5,7 +5,7 @@ tags: [versioning, knowledge, updates, lifecycle]
 created: 2026-07-26
 modified: 2026-08-29
 related:
-  - see-also: Memory Lifecycle
+  - part-of: Memory Lifecycle
  - rests-on: Confidence Scoring
  - see-also: Knowledge Graph
  - exemplifies: LLM Wiki Pattern
@@ -138,7 +138,7 @@ Wenn Aussage B Aussage A ersetzt:
 <!-- wikitool:links -->
 ## Beziehungen

- **see-also:** [[Memory Lifecycle]]
+- **part-of:** [[Memory Lifecycle]]
 - **rests-on:** [[Confidence Scoring]]
 - **see-also:** [[Knowledge Graph]]
 - **exemplifies:** [[LLM Wiki Pattern]]
@@ -6,7 +6,7 @@ created: 2026-08-02
 modified: 2026-08-29
 related:
  - exemplifies: Implementation Spectrum
-  - see-also: Knowledge Graph
+  - part-of: Knowledge Graph
  - evidenced-by: Source - LLM Wiki v2
 sources: []
 confidence: 0.50
@@ -46,6 +46,6 @@ TODO
 ## Beziehungen

 - **exemplifies:** [[Implementation Spectrum]]
- **see-also:** [[Knowledge Graph]]
+- **part-of:** [[Knowledge Graph]]
 - **evidenced-by:** [[Source - LLM Wiki v2]]
 <!-- /wikitool:links -->
@@ -5,7 +5,7 @@ tags: []
 created: 2026-08-02
 modified: 2026-08-29
 related:
-  - see-also: Hybrid Search
+  - part-of: Hybrid Search
  - exemplifies: LLM Wiki Pattern
  - evidenced-by: Source - LLM Wiki v2
 sources: []
@@ -45,7 +45,7 @@ TODO
 <!-- wikitool:links -->
 ## Beziehungen

- **see-also:** [[Hybrid Search]]
+- **part-of:** [[Hybrid Search]]
 - **exemplifies:** [[LLM Wiki Pattern]]
 - **evidenced-by:** [[Source - LLM Wiki v2]]
 <!-- /wikitool:links -->
@@ -6,7 +6,7 @@ created: 2026-08-02
 modified: 2026-08-29
 related:
  - exemplifies: Implementation Spectrum
-  - see-also: Multi-Agent Collaboration
+  - part-of: Multi-Agent Collaboration
 sources: []
 confidence: 0.50
 confidence_base: 0.50
@@ -45,5 +45,5 @@ TODO
 ## Beziehungen

 - **exemplifies:** [[Implementation Spectrum]]
- **see-also:** [[Multi-Agent Collaboration]]
+- **part-of:** [[Multi-Agent Collaboration]]
 <!-- /wikitool:links -->
@@ -5,7 +5,7 @@ tags: []
 created: 2026-08-02
 modified: 2026-08-29
 related:
-  - see-also: Consolidation Tiers
+  - part-of: Consolidation Tiers
 sources: []
 confidence: 0.50
 confidence_base: 0.50
@@ -43,5 +43,5 @@ TODO
 <!-- wikitool:links -->
 ## Beziehungen

- **see-also:** [[Consolidation Tiers]]
+- **part-of:** [[Consolidation Tiers]]
 <!-- /wikitool:links -->
@@ -7,7 +7,7 @@ modified: 2026-08-29
 related:
  - depends-on: Wine
  - see-also: Proton
-  - see-also: Wine GE
+  - part-of: Wine GE
  - see-also: Arch Linux
 sources: [Source - Wine]
 confidence: 0.85
@@ -68,6 +68,6 @@ Wine-Staging-Patches enthalten typischerweise:

 - **depends-on:** [[Wine]]
 - **see-also:** [[Proton]]
- **see-also:** [[Wine GE]]
+- **part-of:** [[Wine GE]]
 - **see-also:** [[Arch Linux]]
 <!-- /wikitool:links -->
@@ -5,7 +5,7 @@ tags: [schema, taxonomy, external, farzaa-gist]
 created: 2026-08-03
 modified: 2026-08-29
 related:
-  - see-also: farzaa gist
+  - part-of: farzaa gist
  - see-also: AGENTS.md
  - evidenced-by: Source - LLM Improvements Sonnet Analysis
 sources: [Source - LLM Improvements Sonnet Analysis]
@@ -79,7 +79,7 @@ Dies sind handlungsfähige Empfehlungen, die in der Sonnet-Analyse als wertvoll
 <!-- wikitool:links -->
 ## Beziehungen

- **see-also:** [[farzaa gist]]
+- **part-of:** [[farzaa gist]]
 - **see-also:** [[AGENTS.md]]
 - **evidenced-by:** [[Source - LLM Improvements Sonnet Analysis]]
 <!-- /wikitool:links -->
@@ -18,7 +18,7 @@ The page tables live in a generated `INDEX.md` inside each collection, linked be
 - **Concepts:** 80
 - **Entities:** 72
 - **Sources:** 28
- **Last Updated:** 2026-09-02
+- **Last Updated:** 2026-09-03

 ---

@@ -127,3 +127,13 @@ Abschluss der Korpus-Migration auf die Link-Taxonomie (Gitea #40, Abschnitte 2 u
 Gitea-Issue #41 als raw/notes/Gitea Issue 41 - Issue Management and Label Scheme 2026-09-02.md aufgenommen und als Source-Seite erfasst. kb/concepts/Issue Label Scheme.md auf den Stand vom 2026-09-02 gebracht: vier Pflicht-Achsen (area/kind/prio/size), zwei optionale status/-Flags, Body-als-Wahrheit-Konvention. Das abgeloeste Zweiachsen-Schema steht als Abschnitt Historie mit Diff-Tabelle in der Seite, nicht geloescht. confidence_base 0.70 -> 0.85 (zweite unabhaengige Quelle, Bestaetigung juenger als 30 Tage).

 ---
+
+## [2026-09-03] update | Link-Taxonomie 4.1.0: 16 see-also-Kanten auf part-of, Comparison-Seite auf gelabelte Kanten
+
+Befund 3 aus #40: composition/part-of ist jetzt das dritte Inversenpaar. Die 16 Gegenkanten eines composition, die im u3-Lauf auf see-also gesetzt wurden, sind per xref add auf part-of relabelt - betroffen sind Consolidation Tiers, Content Quality Control, Hybrid Search, Knowledge Graph, LLM Wiki Pattern, Memory Lifecycle, Multi-Agent Collaboration, Wine GE und farzaa gist samt ihrer Kinder.
+
+Befund 1 aus #40: die einzige Comparison-Seite (amd-pstate vs acpi-cpufreq) trug ihre compares-with-Bullets als handgeschriebene Prosa ohne Frontmatter-Deckung, weil types/comparison.md kein related: fuehrte. Der Type-Spec hat es jetzt; die beiden Kanten stehen als deklarierte Kanten in einer wikitool:links-Region. kb/sources/COLLECTION.md hat seinen inerten outbound:-Block verloren.
+
+migrate verify --from HEAD: 181 Seiten, 0 hinzugefuegt, 0 entfernt, 18 Befunde - alle Label-Wechsel im Frontmatter, keine Aenderung an Wikilink- oder Zitatzahlen. lint --fail-on-error gruen.
+
+---
@@ -1,7 +1,5 @@
 ---
 profile: sources
-outbound:
-  any: [is-evidence-for, defined-in, see-also]
 required_by_stack: true
 ---

@@ -40,16 +38,17 @@ The `raw_files:`/`source_url:`/citation rules are shared and live in
 - `tools/wikitool sources trace --raw <path>` answers "what did we learn from this?";
  `tools/wikitool sources coverage` lists raw files no source page claims yet.

-## Authorised labels
+## No authorised labels

-The `outbound:` block above is what `wikitool lint` and `xref add` check: which labels a page in
-this collection may use, per destination. The catalogue they are drawn from - and what each one
-asserts - is [instructions/link-taxonomy.md](../../instructions/link-taxonomy.md), which binds
-nothing on its own.
+This collection has **no `outbound:` block**, and that is the declaration rather than an
+omission: the `source` type-spec offers no `related:` field, so a source page has nowhere to
+put a labelled edge. Everything it would want to assert is already carried by `raw_files:`,
+`entities:`, `concepts:` and `[^cite-id]` - the mechanical provenance path, not authored edges.

-Deliberately narrow. A source page is evidence *about* a source; almost everything it would want to say is already carried by `raw_files:`, `sources:` and `[^cite-id]`, which are the mechanical provenance path rather than authored edges.
-
-Adding a label here is a deliberate contract change, not a way around a refusal.
+An `outbound:` block here would authorise labels that no page in this collection can write.
+`wikitool docs verify` refuses that combination, so the two cannot drift apart: giving source
+pages labelled edges means giving the type-spec a `related:` field first, which is a deliberate
+contract change and not a way around a refusal.

 ## Outbound linking

@@ -50,7 +50,7 @@ tools/wikitool <command> --help
 | `index rebuild [--dry-run]` | Regenerate the catalog from every page's frontmatter: `kb/index.md` becomes a map (statistics, one row per collection and per area, links to the shards) and the page tables are written to a generated `INDEX.md` in each collection. An area past 50 rows gets its own shard. Stale shards from removed collections/areas are deleted in the same pass |
 | `log append --op ingest\|query\|lint\|create\|update\|delete\|rename --title "..." [--body "..."\|--body-file path]` | Append a formatted entry to `kb/log.md` |
 | `log status` | Read-only: count `ingest` entries logged since the last `lint` entry - the deterministic trigger behind the Maintenance Schedule's "every 10 sources" full-lint cadence |
-| `lint [--json] [--markdown out.md] [--full] [--fail-on-error]` | Structural + provenance checks: broken wikilinks, dangling frontmatter references, orphan pages, index drift, schema gaps, duplicate titles, title mismatches, uncovered raw files, broken `raw_files:` refs, raw files claimed by more than one source page, unmarked provenance, citation/frontmatter drift, quote-limit overages (>2 blockquoted lines/page, advisory only). Prints only the sections that found something and always writes the full report to `reports/Lint Report <date>.md` (or `--markdown`), naming the path - `--full` prints everything, `--json` prints the findings and writes nothing |
+| `lint [--json] [--markdown out.md] [--full] [--fail-on-error]` | Structural + provenance checks: broken wikilinks, dangling frontmatter references, orphan pages, index drift, schema gaps, duplicate titles, title mismatches, uncovered raw files, broken `raw_files:` refs, raw files claimed by more than one source page, unmarked provenance, citation/frontmatter drift, unbalanced generated-region markers, edges whose label is missing or not authorised by the source collection's `outbound:` (both hard once `kb_version` has reached the release that introduced labelled edges - advisory below it, so a corpus mid-migration is not refused by the check measuring it), quote-limit overages (>2 blockquoted lines/page, advisory only). Prints only the sections that found something and always writes the full report to `reports/Lint Report <date>.md` (or `--markdown`), naming the path - `--full` prints everything, `--json` prints the findings and writes nothing |
 | `search ["<text>"] [--field <predicate> ...] [--kind/--subtype/--collection/--tag <v>] [--regex] [--limit N] [--sort [-]<field>] [--backend <name>] [--matches] [--json]` | Find pages in `kb/` without reading the index. Text search runs through a pluggable backend (`rg` today); `--field` predicates are evaluated on frontmatter - `f=v`, `f~substring`, `'f>=v'`, `'f:*'` (present), `'!f'` (absent), repeatable and ANDed. With no text this is a pure structured query. Results carry kind/summary/confidence so a hit can be judged without opening the page. A page whose frontmatter does not parse can match no positive predicate, so it is **named** rather than dropped: `--json` always carries an `unreadable` list of `{path, reason}` (usually empty), and the table form writes the same lines to stderr. `--regex` is applied by `rg` alone, whose engine is linear; the ranking boosts for title and summary are literal-containment only, so a non-literal pattern is ranked by match count. `rg` is killed after 30 s and reported as a failure. Read-only, and **exempt from the Iteration Budget Gate** |
 | `confidence decay [--apply]` | Recompute every page's derived `confidence` as `confidence_base * (1 - 0.01/month)`, floored at 0.2; dry-run by default |
 | `confidence init-base [--apply]` | One-time backfill: set `confidence_base` from the current `confidence` on pages that predate the derived-confidence model |
@@ -16,10 +16,12 @@ from chemenu import config
 from chemenu.commands._util import rel_path, success
 from chemenu.lint_core import (
    HARD_ERROR_KEYS,
+    MIGRATION_GATED_KEYS,
    MOST_LINKED_COUNT,
    QUOTE_LIMIT,
    count_quote_blocks,
    default_report_path,
+    hard_error_keys,
    has_hard_errors,
    render_markdown,
    render_summary,
@@ -30,10 +32,12 @@ from chemenu.lint_core import (
 # so does every other name the tests and sibling commands already import.
 __all__ = [
    "HARD_ERROR_KEYS",
+    "MIGRATION_GATED_KEYS",
    "MOST_LINKED_COUNT",
    "QUOTE_LIMIT",
    "count_quote_blocks",
    "default_report_path",
+    "hard_error_keys",
    "has_hard_errors",
    "render_markdown",
    "render_summary",
@@ -183,6 +183,33 @@ def authorised_labels(source: str, destination: str, kb_dir: Path | None = None)
    return labels


+LABELLED_EDGE_FIELD = "related"
+
+
+def collections_that_can_carry_labelled_edges() -> set[str]:
+    """Collection names whose offered page types actually have somewhere to put
+    a labelled edge.
+
+    Derived from `page_ref_fields:`, the same way `stack_required_collections()`
+    is derived from `base_dir:`: a type that does not offer `related:` cannot
+    carry a label, no matter what its collection's contract authorises.
+    """
+    from chemenu.type_resolver import resolver
+
+    names: set[str] = set()
+    for type_path, _frontmatter in resolver.list_type_specs():
+        try:
+            if resolver.get_root(type_path) != "kb":
+                continue
+            base_dir = resolver.get_base_dir(type_path)
+            fields = resolver.get_page_ref_fields(type_path)
+        except (ValueError, OSError):
+            continue
+        if base_dir and LABELLED_EDGE_FIELD in fields:
+            names.add(str(base_dir).strip("/"))
+    return names
+
+
 def declaration_issues(kb_dir: Path | None = None) -> list[str]:
    """What each `COLLECTION.md` fails to declare about itself.

@@ -200,6 +227,7 @@ def declaration_issues(kb_dir: Path | None = None) -> list[str]:
    issues: list[str] = []

    required = stack_required_collections()
+    can_label = collections_that_can_carry_labelled_edges()
    present = {path.name for path in iter_kb_collections(root)}
    for name in required:
        if name not in present:
@@ -244,6 +272,20 @@ def declaration_issues(kb_dir: Path | None = None) -> list[str]:
                )
                + f" - it must be {str(expected).lower()}"
            )
+
+        # An `outbound:` block on a collection whose types offer no `related:`
+        # authorises labels that no page there can write. That is not a harmless
+        # extra: it reads as a licence, so the label gets written into the prose
+        # by hand instead - an identifier back in free text, which is the exact
+        # thing the labelled-edge model exists to end. The two halves have to
+        # move together, so the check names both directions of the fix.
+        if declared.get(OUTBOUND_FIELD) and collection.name not in can_label:
+            issues.append(
+                f"{relative}: `{OUTBOUND_FIELD}:` authorises labels, but no page type writing "
+                f"into kb/{collection.name}/ offers a `{LABELLED_EDGE_FIELD}:` field - so no "
+                f"page here can carry a labelled edge. Either drop the block, or give the "
+                f"type-spec a `{LABELLED_EDGE_FIELD}:` in its `page_ref_fields:` and schema"
+            )
    return issues


@@ -27,6 +27,7 @@ from chemenu.provenance import legacy_source_pages as find_legacy_source_pages
 from chemenu.provenance import orphan_footnote_defs as find_orphan_footnote_defs
 from chemenu.provenance import uncovered_raw_files as find_uncovered_raw_files
 from chemenu.provenance import undefined_footnote_refs as find_undefined_footnote_refs
+from chemenu.version import Version
 from chemenu.kb_scan import (
    GENERATED_INDEX,
    WIKILINK_RE,
@@ -478,13 +479,6 @@ def default_report_path(report: dict) -> Path:
 # through the index or navigation only. `quote_limit_violations` is advisory
 # too - it flags a habit, not a broken tree.
 #
-# `unlabelled_edges` and `unauthorised_labels` are advisory **for now**, and
-# that is a dated decision rather than a judgment about severity: they describe
-# exactly the state a corpus is in between the 4.0.0 machinery landing and the
-# migration reaching each page, which is the window `.wikitool-kb.json` exists
-# to represent. They become hard errors once the migration is recorded - the
-# same path `legacy_citation_markers` took.
-#
 # `malformed_edges` and `unbalanced_markers` are hard from the start: neither
 # describes an unconverted page, only a broken one.
 #
@@ -505,11 +499,57 @@ HARD_ERROR_KEYS = (
    "dangling_frontmatter_refs",
    "malformed_edges",
    "unbalanced_markers",
+    "unlabelled_edges",
+    "unauthorised_labels",
    "invalid_type_paths",
    "type_resolution_errors",
    "schema_validation_errors",
 )

+# Findings that only become hard once the corpus has reached a given shape.
+#
+# `unlabelled_edges` and `unauthorised_labels` describe exactly the state a
+# corpus is in between the 4.0.0 machinery landing and the migration reaching
+# each page - the window `.wikitool-kb.json` exists to represent. Failing on
+# them during that window would refuse the very corpus that
+# `instructions/migrations/4.0.0-link-taxonomy.md` tells an instance to publish
+# unit by unit. So the promotion is tied to `kb_version` rather than to a
+# release date: below 4.0.0 they are advisory, at or above it an unlabelled
+# edge is no longer a page awaiting conversion but an edge whose author did not
+# say what it asserts.
+#
+# Gated rather than simply promoted, which is where this departs from
+# `legacy_citation_markers`: that one was flipped in a later version and any
+# instance still owing the citation migration had to live with a red lint. The
+# ledger can answer the question now, so it does.
+MIGRATION_GATED_KEYS: dict[str, Version] = {
+    "unlabelled_edges": Version(4, 0, 0),
+    "unauthorised_labels": Version(4, 0, 0),
+}
+
+_ALWAYS_HARD = Version(0, 0, 0)
+
+
+def hard_error_keys(kb_version: Version | None = None) -> tuple[str, ...]:
+    """`HARD_ERROR_KEYS` minus the findings this corpus has not grown into yet.
+
+    `kb_version` defaults to what `.wikitool-kb.json` records. A tree without
+    one - a fresh instance, which starts at the current shape rather than
+    migrating into it - keeps every key: there is no outstanding migration for
+    a gated finding to be the noise of.
+    """
+    from chemenu import kb_state
+
+    if kb_version is None:
+        kb_version = kb_state.read_kb_version()
+    if kb_version is None:
+        return HARD_ERROR_KEYS
+    return tuple(
+        key
+        for key in HARD_ERROR_KEYS
+        if kb_version >= MIGRATION_GATED_KEYS.get(key, _ALWAYS_HARD)
+    )
+

 def has_hard_errors(report: dict) -> bool:
-    return any(report.get(key) for key in HARD_ERROR_KEYS)
+    return any(report.get(key) for key in hard_error_keys())
@@ -192,3 +192,29 @@ def test_an_undeclared_destination_authorises_nothing(kb_root):
    missing declaration to be filled in with a permissive default."""
    _authorising(kb_root, "entities", "  concepts: [implements]")
    assert kb_collections.authorised_labels("entities", "sources") == set()
+
+
+def test_outbound_on_a_collection_that_cannot_carry_labels_is_a_finding(kb_root):
+    """`kb/sources/` is the live case: the `source` type-spec offers no
+    `related:`, so an `outbound:` block there authorises labels no page can
+    write. Left unchecked it reads as a licence and the label gets written into
+    the prose by hand instead - an identifier back in free text, which is what
+    labelled edges exist to end."""
+    _authorising(kb_root, "sources", "  any: [is-evidence-for]", required=True)
+    issues = kb_collections.declaration_issues(kb_root)
+    assert any(
+        "kb/sources/COLLECTION.md" in issue and "no page type writing into kb/sources/" in issue
+        for issue in issues
+    )
+
+
+def test_outbound_is_fine_on_a_collection_whose_type_offers_related(kb_root):
+    _collection(kb_root, "sources", profile="sources", required=True)
+    _authorising(kb_root, "entities", "  any: [uses]")
+    assert kb_collections.declaration_issues(kb_root) == []
+
+
+def test_a_collection_without_outbound_is_not_a_finding(kb_root):
+    """Absence is the declaration `kb/sources/` makes: no authored edges here."""
+    _collection(kb_root, "sources", profile="sources", required=True)
+    assert kb_collections.declaration_issues(kb_root) == []
@@ -1,8 +1,12 @@
 import json
 from datetime import date

-from chemenu import config
+import pytest
+
+from chemenu import config, kb_state
 from chemenu.commands.lint import (
+    HARD_ERROR_KEYS,
+    hard_error_keys,
    has_hard_errors,
    lint_command,
    render_markdown,
@@ -11,6 +15,8 @@ from chemenu.commands.lint import (
 )
 from chemenu.frontmatter_io import write_page
 from chemenu.provenance import cite_id, render_cite_block
+from chemenu.tests.conftest import use_shipped_type_specs
+from chemenu.version import Version


 def test_lint_detects_unparsable_frontmatter(kb_dir):
@@ -472,3 +478,83 @@ def test_lint_does_not_count_a_shell_prompt_as_a_quote(kb_dir):
    )
    report = run_lint(kb_dir)
    assert [i for i in report["quote_limit_violations"] if i["page"] == "shelly"] == []
+
+
+# --- the migration gate on `unlabelled_edges` / `unauthorised_labels` -------
+#
+# These four repoint `config.ROOT` at the fixture tree, which the `kb_dir`
+# fixture alone does not do. Two things need it: `write_kb_state()` writes
+# `.wikitool-kb.json` relative to `ROOT`, and `lint`'s collection lookup
+# resolves a page against `config.KB_DIR` rather than against the directory it
+# was handed - so without the repoint the gate would be read off the real
+# repository's state file.
+
+
+@pytest.fixture
+def rooted_kb(kb_dir, tmp_path, monkeypatch):
+    monkeypatch.setattr(config, "ROOT", tmp_path)
+    use_shipped_type_specs(monkeypatch)
+    return kb_dir
+
+
+def _page_with_an_unlabelled_edge(rooted_kb):
+    """A `related:` entry that is a bare title rather than a `label: title`
+    mapping - the shape every page was in before the 4.0.0 migration."""
+    write_page(
+        rooted_kb / "entities/tools/bare-edge.md",
+        {"type": "types/entity.md", "entity_type": "tool", "tags": [], "created": "2026-09-03",
+         "modified": "2026-09-03", "related": ["Modbus"], "sources": [], "confidence": 0.8,
+         "provenance": "general", "summary": "One edge whose label was never declared."},
+        "\n# bare-edge\n\nAn edge without a label.\n",
+    )
+
+
+def test_unlabelled_edge_is_advisory_below_kb_version_4(rooted_kb):
+    """The window the migration document describes: the machinery has landed,
+    the corpus has not been converted yet, and `lint --fail-on-error` must not
+    refuse the very tree the migration tells the instance to publish unit by
+    unit."""
+    _page_with_an_unlabelled_edge(rooted_kb)
+    kb_state.write_kb_state(Version(3, 0, 0), [])
+    report = run_lint(rooted_kb)
+    assert report["unlabelled_edges"] != []
+    assert "unlabelled_edges" not in hard_error_keys()
+    # Narrowed to the finding under test: the fixture corpus carries unrelated
+    # hard errors of its own, so asserting on the whole report would prove
+    # nothing about the gate.
+    assert has_hard_errors({"unlabelled_edges": report["unlabelled_edges"]}) is False
+
+
+def test_unlabelled_edge_is_hard_at_kb_version_4(rooted_kb):
+    """Once the migration is recorded, a bare title is no longer a page waiting
+    its turn - it is an edge whose author did not say what it asserts."""
+    _page_with_an_unlabelled_edge(rooted_kb)
+    kb_state.write_kb_state(Version(4, 0, 0), [])
+    report = run_lint(rooted_kb)
+    assert report["unlabelled_edges"] != []
+    assert "unlabelled_edges" in hard_error_keys()
+    assert has_hard_errors({"unlabelled_edges": report["unlabelled_edges"]}) is True
+
+
+def test_unauthorised_label_is_hard_at_kb_version_4(rooted_kb):
+    """The fixture contracts authorise `depends-on` but not `contradicts`."""
+    write_page(
+        rooted_kb / "entities/tools/off-menu.md",
+        {"type": "types/entity.md", "entity_type": "tool", "tags": [], "created": "2026-09-03",
+         "modified": "2026-09-03", "related": [{"contradicts": "Modbus"}], "sources": [],
+         "confidence": 0.8, "provenance": "general", "summary": "A label off this menu."},
+        "\n# off-menu\n\nA label the source collection never authorised.\n",
+    )
+    kb_state.write_kb_state(Version(4, 0, 0), [])
+    report = run_lint(rooted_kb)
+    assert report["unauthorised_labels"] != []
+    assert "unauthorised_labels" in hard_error_keys()
+    assert has_hard_errors({"unauthorised_labels": report["unauthorised_labels"]}) is True
+
+
+def test_a_tree_that_never_declared_a_kb_version_keeps_every_key(rooted_kb):
+    """No `.wikitool-kb.json` means a fresh instance, which starts at the
+    current shape rather than migrating into it - so there is no outstanding
+    migration for a gated finding to be the noise of."""
+    assert kb_state.read_kb_version() is None
+    assert hard_error_keys() == HARD_ERROR_KEYS
@@ -45,7 +45,7 @@ def test_get_page_ref_fields_reads_the_type_spec():
    assert resolver.get_page_ref_fields("types/entity.md") == ["related", "sources"]
    assert resolver.get_page_ref_fields("types/concept.md") == ["related", "sources"]
    assert resolver.get_page_ref_fields("types/source.md") == ["entities", "concepts"]
-    assert resolver.get_page_ref_fields("types/comparison.md") == ["entities"]
+    assert resolver.get_page_ref_fields("types/comparison.md") == ["entities", "related"]


 def test_page_ref_fields_exist_in_the_type_schema():
@@ -4,7 +4,7 @@ name: comparison
 description: Strukturierter Typ für Vergleichsseiten, die mehrere Entities oder Ansätze gegenüberstellen
 schema: types/comparison.schema.yaml
 base_dir: comparisons
-page_ref_fields: [entities]
+page_ref_fields: [entities, related]
 ---

 # Comparison
@@ -32,6 +32,7 @@ page_ref_fields: [entities]
 | `tags` | Nein | Navigations-Tags zur Kategorisierung |
 | `created` | Ja | Erstellungsdatum (YYYY-MM-DD) |
 | `entities` | Ja | Titel der verglichenen Entities |
+| `related` | Nein | Deklarierte ausgehende Kanten - je Subjekt eine `compares-with`-Kante, geschrieben von `wikitool xref add` |
 | `summary` | Ja | Einzeiler für `kb/index.md` |

 ## Autorenanweisungen
@@ -21,6 +21,23 @@ properties:
      type: string
    description: Entity titles being compared
    minItems: 2
+  related:
+    type: array
+    items:
+      oneOf:
+        - type: string
+        - type: object
+          minProperties: 1
+          maxProperties: 1
+          additionalProperties:
+            type: string
+    description: >-
+      Declared outbound edges, in the same shape entity and concept pages use.
+      A comparison's own assertion is `compares-with` against each subject: the
+      titles are already in `entities:`, but that field is the untyped
+      provenance-style list, so without this one the edge the page exists to
+      make would live only in hand-written prose - which is the thing
+      instructions/link-taxonomy.md was built to end.
  summary:
    type: string
    description: 1-line summary for index.md
@@ -1 +1 @@
 .0.1
 .1.0