@wildo-ai/framework-knowledge 1.1.1 → 1.1.3

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (223) hide show
  1. package/dist/esm/framework-knowledge.definitions.d.ts +1 -1
  2. package/dist/esm/framework-knowledge.definitions.d.ts.map +1 -1
  3. package/dist/esm/framework-knowledge.definitions.js +271 -8
  4. package/dist/esm/framework-knowledge.definitions.js.map +1 -1
  5. package/dist/esm/framework-knowledge.loader.d.ts +19 -0
  6. package/dist/esm/framework-knowledge.loader.d.ts.map +1 -1
  7. package/dist/esm/framework-knowledge.loader.js +32 -0
  8. package/dist/esm/framework-knowledge.loader.js.map +1 -1
  9. package/dist/esm/framework-knowledge.schemas.d.ts +13 -2
  10. package/dist/esm/framework-knowledge.schemas.d.ts.map +1 -1
  11. package/dist/esm/framework-knowledge.schemas.js.map +1 -1
  12. package/dist/esm/playbooks/bottom_up.assemble-verify/playbook.definition.d.ts +45 -0
  13. package/dist/esm/playbooks/bottom_up.assemble-verify/playbook.definition.d.ts.map +1 -0
  14. package/dist/esm/playbooks/bottom_up.assemble-verify/playbook.definition.js +55 -0
  15. package/dist/esm/playbooks/bottom_up.assemble-verify/playbook.definition.js.map +1 -0
  16. package/dist/esm/playbooks/bottom_up.build-decompose/playbook.definition.d.ts +6 -7
  17. package/dist/esm/playbooks/bottom_up.build-decompose/playbook.definition.d.ts.map +1 -1
  18. package/dist/esm/playbooks/bottom_up.build-decompose/playbook.definition.js +6 -6
  19. package/dist/esm/playbooks/bottom_up.build-decompose/playbook.definition.js.map +1 -1
  20. package/dist/esm/playbooks/bottom_up.build-dispatch/playbook.definition.d.ts +0 -4
  21. package/dist/esm/playbooks/bottom_up.build-dispatch/playbook.definition.d.ts.map +1 -1
  22. package/dist/esm/playbooks/bottom_up.build-dispatch/playbook.definition.js +0 -2
  23. package/dist/esm/playbooks/bottom_up.build-dispatch/playbook.definition.js.map +1 -1
  24. package/dist/esm/playbooks/bottom_up.build-scale/playbook.definition.d.ts +0 -4
  25. package/dist/esm/playbooks/bottom_up.build-scale/playbook.definition.d.ts.map +1 -1
  26. package/dist/esm/playbooks/bottom_up.build-scale/playbook.definition.js +0 -2
  27. package/dist/esm/playbooks/bottom_up.build-scale/playbook.definition.js.map +1 -1
  28. package/dist/esm/playbooks/bottom_up.domain-plan/playbook.definition.d.ts +3 -9
  29. package/dist/esm/playbooks/bottom_up.domain-plan/playbook.definition.d.ts.map +1 -1
  30. package/dist/esm/playbooks/bottom_up.domain-plan/playbook.definition.js +1 -3
  31. package/dist/esm/playbooks/bottom_up.domain-plan/playbook.definition.js.map +1 -1
  32. package/dist/esm/playbooks/bottom_up.foundation/playbook.definition.d.ts +10 -6
  33. package/dist/esm/playbooks/bottom_up.foundation/playbook.definition.d.ts.map +1 -1
  34. package/dist/esm/playbooks/bottom_up.foundation/playbook.definition.js +25 -4
  35. package/dist/esm/playbooks/bottom_up.foundation/playbook.definition.js.map +1 -1
  36. package/dist/esm/playbooks/bottom_up.placement-plan/playbook.definition.d.ts +1 -5
  37. package/dist/esm/playbooks/bottom_up.placement-plan/playbook.definition.d.ts.map +1 -1
  38. package/dist/esm/playbooks/bottom_up.placement-plan/playbook.definition.js +1 -3
  39. package/dist/esm/playbooks/bottom_up.placement-plan/playbook.definition.js.map +1 -1
  40. package/dist/esm/playbooks/refinement.compliance-document-candidate/playbook.definition.d.ts +4 -2
  41. package/dist/esm/playbooks/refinement.compliance-document-candidate/playbook.definition.d.ts.map +1 -1
  42. package/dist/esm/playbooks/refinement.compliance-document-candidate/playbook.definition.js +4 -2
  43. package/dist/esm/playbooks/refinement.compliance-document-candidate/playbook.definition.js.map +1 -1
  44. package/dist/esm/playbooks/refinement.document-facts/playbook.definition.d.ts +1 -2
  45. package/dist/esm/playbooks/refinement.document-facts/playbook.definition.d.ts.map +1 -1
  46. package/dist/esm/playbooks/refinement.document-facts/playbook.definition.js +13 -14
  47. package/dist/esm/playbooks/refinement.document-facts/playbook.definition.js.map +1 -1
  48. package/dist/esm/playbooks/top_down.brand-identity/playbook.definition.d.ts +3 -5
  49. package/dist/esm/playbooks/top_down.brand-identity/playbook.definition.d.ts.map +1 -1
  50. package/dist/esm/playbooks/top_down.brand-identity/playbook.definition.js +3 -4
  51. package/dist/esm/playbooks/top_down.brand-identity/playbook.definition.js.map +1 -1
  52. package/dist/esm/playbooks/top_down.business-model/playbook.definition.d.ts +2 -8
  53. package/dist/esm/playbooks/top_down.business-model/playbook.definition.d.ts.map +1 -1
  54. package/dist/esm/playbooks/top_down.business-model/playbook.definition.js +0 -2
  55. package/dist/esm/playbooks/top_down.business-model/playbook.definition.js.map +1 -1
  56. package/dist/esm/playbooks/top_down.capability-map/playbook.definition.d.ts +2 -8
  57. package/dist/esm/playbooks/top_down.capability-map/playbook.definition.d.ts.map +1 -1
  58. package/dist/esm/playbooks/top_down.capability-map/playbook.definition.js +0 -2
  59. package/dist/esm/playbooks/top_down.capability-map/playbook.definition.js.map +1 -1
  60. package/dist/esm/playbooks/top_down.compliance-analysis/playbook.definition.d.ts +52 -0
  61. package/dist/esm/playbooks/top_down.compliance-analysis/playbook.definition.d.ts.map +1 -0
  62. package/dist/esm/playbooks/top_down.compliance-analysis/playbook.definition.js +127 -0
  63. package/dist/esm/playbooks/top_down.compliance-analysis/playbook.definition.js.map +1 -0
  64. package/dist/esm/playbooks/top_down.compliance-conception/playbook.definition.d.ts +75 -0
  65. package/dist/esm/playbooks/top_down.compliance-conception/playbook.definition.d.ts.map +1 -0
  66. package/dist/esm/playbooks/top_down.compliance-conception/playbook.definition.js +91 -0
  67. package/dist/esm/playbooks/top_down.compliance-conception/playbook.definition.js.map +1 -0
  68. package/dist/esm/playbooks/top_down.compliance-evidence-adequacy/playbook.definition.d.ts +29 -0
  69. package/dist/esm/playbooks/top_down.compliance-evidence-adequacy/playbook.definition.d.ts.map +1 -0
  70. package/dist/esm/playbooks/top_down.compliance-evidence-adequacy/playbook.definition.js +28 -0
  71. package/dist/esm/playbooks/top_down.compliance-evidence-adequacy/playbook.definition.js.map +1 -0
  72. package/dist/esm/playbooks/top_down.compliance-posture/playbook.definition.d.ts +9 -45
  73. package/dist/esm/playbooks/top_down.compliance-posture/playbook.definition.d.ts.map +1 -1
  74. package/dist/esm/playbooks/top_down.compliance-posture/playbook.definition.js +4 -54
  75. package/dist/esm/playbooks/top_down.compliance-posture/playbook.definition.js.map +1 -1
  76. package/dist/esm/playbooks/top_down.compliance-research/playbook.definition.d.ts +30 -0
  77. package/dist/esm/playbooks/top_down.compliance-research/playbook.definition.d.ts.map +1 -0
  78. package/dist/esm/playbooks/top_down.compliance-research/playbook.definition.js +25 -0
  79. package/dist/esm/playbooks/top_down.compliance-research/playbook.definition.js.map +1 -0
  80. package/dist/esm/playbooks/top_down.market-analysis/playbook.definition.d.ts +2 -9
  81. package/dist/esm/playbooks/top_down.market-analysis/playbook.definition.d.ts.map +1 -1
  82. package/dist/esm/playbooks/top_down.market-analysis/playbook.definition.js +9 -5
  83. package/dist/esm/playbooks/top_down.market-analysis/playbook.definition.js.map +1 -1
  84. package/dist/esm/playbooks/top_down.market-evidence-adequacy/playbook.definition.d.ts +0 -1
  85. package/dist/esm/playbooks/top_down.market-evidence-adequacy/playbook.definition.d.ts.map +1 -1
  86. package/dist/esm/playbooks/top_down.market-evidence-adequacy/playbook.definition.js +0 -1
  87. package/dist/esm/playbooks/top_down.market-evidence-adequacy/playbook.definition.js.map +1 -1
  88. package/dist/esm/playbooks/top_down.market-research/playbook.definition.d.ts +0 -1
  89. package/dist/esm/playbooks/top_down.market-research/playbook.definition.d.ts.map +1 -1
  90. package/dist/esm/playbooks/top_down.market-research/playbook.definition.js +0 -1
  91. package/dist/esm/playbooks/top_down.market-research/playbook.definition.js.map +1 -1
  92. package/dist/esm/playbooks/top_down.market-snapshot-produce/playbook.definition.d.ts +0 -9
  93. package/dist/esm/playbooks/top_down.market-snapshot-produce/playbook.definition.d.ts.map +1 -1
  94. package/dist/esm/playbooks/top_down.market-snapshot-produce/playbook.definition.js +0 -2
  95. package/dist/esm/playbooks/top_down.market-snapshot-produce/playbook.definition.js.map +1 -1
  96. package/dist/esm/playbooks/top_down.phase-gate/playbook.definition.d.ts +2 -8
  97. package/dist/esm/playbooks/top_down.phase-gate/playbook.definition.d.ts.map +1 -1
  98. package/dist/esm/playbooks/top_down.phase-gate/playbook.definition.js +0 -2
  99. package/dist/esm/playbooks/top_down.phase-gate/playbook.definition.js.map +1 -1
  100. package/dist/esm/playbooks/top_down.requirements-definition/playbook.definition.d.ts +2 -8
  101. package/dist/esm/playbooks/top_down.requirements-definition/playbook.definition.d.ts.map +1 -1
  102. package/dist/esm/playbooks/top_down.requirements-definition/playbook.definition.js +0 -2
  103. package/dist/esm/playbooks/top_down.requirements-definition/playbook.definition.js.map +1 -1
  104. package/dist/esm/playbooks/top_down.scope-roadmap/playbook.definition.d.ts +2 -8
  105. package/dist/esm/playbooks/top_down.scope-roadmap/playbook.definition.d.ts.map +1 -1
  106. package/dist/esm/playbooks/top_down.scope-roadmap/playbook.definition.js +5 -4
  107. package/dist/esm/playbooks/top_down.scope-roadmap/playbook.definition.js.map +1 -1
  108. package/dist/esm/playbooks/top_down.value-proposition/playbook.definition.d.ts +0 -4
  109. package/dist/esm/playbooks/top_down.value-proposition/playbook.definition.d.ts.map +1 -1
  110. package/dist/esm/playbooks/top_down.value-proposition/playbook.definition.js +0 -2
  111. package/dist/esm/playbooks/top_down.value-proposition/playbook.definition.js.map +1 -1
  112. package/dist/esm/playbooks/top_down.vision-elicitation/playbook.definition.d.ts +4 -6
  113. package/dist/esm/playbooks/top_down.vision-elicitation/playbook.definition.d.ts.map +1 -1
  114. package/dist/esm/playbooks/top_down.vision-elicitation/playbook.definition.js +25 -5
  115. package/dist/esm/playbooks/top_down.vision-elicitation/playbook.definition.js.map +1 -1
  116. package/dist/knowledge/manifest.json +6 -7
  117. package/dist/knowledge/playbooks/bottom_up.build-decompose/method.md +7 -47
  118. package/dist/knowledge/playbooks/bottom_up.build-decompose/playbook.json +6 -13
  119. package/dist/knowledge/playbooks/bottom_up.build-dispatch/method.md +1 -82
  120. package/dist/knowledge/playbooks/bottom_up.build-dispatch/playbook.json +11 -6
  121. package/dist/knowledge/playbooks/bottom_up.build-scale/playbook.json +10 -5
  122. package/dist/knowledge/playbooks/bottom_up.domain-plan/playbook.json +3 -4
  123. package/dist/knowledge/playbooks/bottom_up.foundation/playbook.json +3 -4
  124. package/dist/knowledge/playbooks/refinement.compliance-document-candidate/method.md +15 -735
  125. package/dist/knowledge/playbooks/refinement.compliance-document-candidate/playbook.json +11 -7
  126. package/dist/knowledge/playbooks/refinement.document-facts/method.md +16 -123
  127. package/dist/knowledge/playbooks/refinement.document-facts/playbook.json +15 -8
  128. package/dist/knowledge/playbooks/top_down.brand-identity/playbook.json +10 -4
  129. package/dist/knowledge/playbooks/top_down.business-model/playbook.json +10 -4
  130. package/dist/knowledge/playbooks/top_down.capability-map/method.md +2 -2
  131. package/dist/knowledge/playbooks/top_down.capability-map/playbook.json +12 -5
  132. package/dist/knowledge/playbooks/top_down.compliance-posture/playbook.json +10 -4
  133. package/dist/knowledge/playbooks/{top_down.market-snapshot-produce → top_down.market-analysis}/evaluation-protocol.json +1 -1
  134. package/dist/knowledge/playbooks/{top_down.market-snapshot-produce → top_down.market-analysis}/evaluation-treatment.json +1 -1
  135. package/dist/knowledge/playbooks/top_down.market-analysis/method.md +165 -51
  136. package/dist/knowledge/playbooks/top_down.market-analysis/playbook.json +39 -21
  137. package/dist/knowledge/playbooks/top_down.phase-gate/playbook.json +2 -6
  138. package/dist/knowledge/playbooks/top_down.requirements-definition/playbook.json +10 -4
  139. package/dist/knowledge/playbooks/top_down.scope-roadmap/method.md +1 -1
  140. package/dist/knowledge/playbooks/top_down.scope-roadmap/playbook.json +11 -5
  141. package/dist/knowledge/playbooks/top_down.value-proposition/method.md +2 -2
  142. package/dist/knowledge/playbooks/top_down.value-proposition/playbook.json +11 -5
  143. package/dist/knowledge/playbooks/top_down.vision-elicitation/playbook.json +10 -4
  144. package/dist/knowledge/references/build-plan.task-graph.config.md +1 -45
  145. package/dist/knowledge/references/example.config.app-shell.full.md +4 -0
  146. package/dist/knowledge/references/example.provider.backend-module-shape.md +0 -7
  147. package/dist/knowledge/references/example.resource.add-resource.module-factory-map.md +1 -0
  148. package/dist/knowledge/references/platform.config.saas.md +0 -50
  149. package/dist/knowledge/references/saas.external-data.admin-endpoints.md +48 -0
  150. package/dist/knowledge/references/saas.frontend.app-level-component.configuration.md +10 -6
  151. package/dist/knowledge/references/saas.frontend.app-level-component.vocabulary.md +1 -0
  152. package/dist/knowledge/references/saas.frontend.resource-operation-frontend-config.md +0 -109
  153. package/dist/knowledge/references/saas.models.data-mode-read-list.md +0 -137
  154. package/dist/knowledge/references/saas.models.http-api-binding.md +29 -3
  155. package/dist/knowledge/references/saas.models.resource-config.builder-parameters.md +2 -4
  156. package/dist/knowledge/references/saas.models.resource-config.resource-configuration.md +0 -2
  157. package/dist/knowledge/references/saas.models.resource-config.resources-registry.md +0 -14
  158. package/dist/knowledge/references/saas.specifications.application-document-fact-specification.contract.md +1 -1
  159. package/dist/knowledge/references/saas.specifications.companion-exports.contract.md +0 -36
  160. package/dist/knowledge/references/zod.decorators.chainable-runtime.md +0 -26
  161. package/dist/knowledge/references/zod.decorators.decorator-vocabulary.md +0 -45
  162. package/dist/knowledge/references/zod.decorators.display-edit-modes.md +1 -1
  163. package/dist/knowledge/references/zod.decorators.per-primitive-ui.md +1 -1
  164. package/dist/knowledge/references/zod.decorators.schema-metadata.md +0 -1
  165. package/dist/knowledge/registries/assurance-authority-bases.json +1 -1
  166. package/dist/knowledge/registries/framework-topics.json +11 -48
  167. package/dist/knowledge/registries/playbooks.json +11 -74
  168. package/dist/knowledge/registries/source-tags.json +538 -622
  169. package/dist/knowledge/skills/INDEX.md +0 -1
  170. package/dist/knowledge/skills/compliance-document-authoring/SKILL.md +2 -16
  171. package/dist/knowledge/skills/compliance-document-authoring/references/document-composition-governance.md +8 -26
  172. package/dist/knowledge/skills/compliance-document-authoring/references/document-composition-privacy-notice.md +74 -252
  173. package/dist/knowledge/skills/compliance-document-authoring/references/document-composition-service-terms.md +8 -26
  174. package/dist/knowledge/skills/compliance-document-authoring/references/framework-application-document-fact-specification-contract.md +21 -36
  175. package/dist/knowledge/skills/compliance-document-authoring/references/framework-assurance-document-blueprint-contract.md +19 -49
  176. package/dist/knowledge/skills/config-app-shell/SKILL.md +8 -6
  177. package/dist/knowledge/skills/config-app-shell/references/framework-app-shell-api.md +11 -6
  178. package/dist/knowledge/skills/config-saas/references/framework-wildo-saas-config-api.md +0 -50
  179. package/dist/knowledge/skills/configure-billing/references/example-backend-provider-module-shape.md +0 -7
  180. package/dist/knowledge/skills/market-companion-prompt-enrichment/SKILL.md +0 -36
  181. package/dist/knowledge/skills/market-companion-prompt-enrichment/references/framework-companion-behavior-api.md +0 -36
  182. package/dist/knowledge/skills/module-companion-exports/references/companion-exports-by-package.md +0 -36
  183. package/dist/knowledge/skills/resource-add-field/references/framework-field-cross-layer-api.md +0 -109
  184. package/dist/knowledge/skills/resource-add-resource/SKILL.md +1 -2
  185. package/dist/knowledge/skills/resource-add-resource/references/framework-new-resource-core-config-api.md +2 -4
  186. package/dist/knowledge/skills/resource-config-wiring/SKILL.md +0 -2
  187. package/dist/knowledge/skills/resource-config-wiring/references/framework-saas-models-resource-config-api.md +2 -157
  188. package/dist/knowledge/skills/resource-custom-view/references/custom-view-contracts.md +0 -109
  189. package/dist/knowledge/skills/resource-external-data/SKILL.md +2 -0
  190. package/dist/knowledge/skills/resource-external-data/references/framework-http-api-binding-api.md +29 -3
  191. package/dist/knowledge/skills/resource-external-data/references/framework-pipeline-admin-endpoints-api.md +70 -0
  192. package/dist/knowledge/skills/resource-schema-authoring/SKILL.md +0 -1
  193. package/dist/knowledge/skills/resource-schema-authoring/references/decorator-layer-map.md +0 -17
  194. package/dist/knowledge/skills/resource-schema-authoring/references/framework-decorator-vocabulary-api.md +0 -46
  195. package/dist/knowledge/skills/resource-schema-authoring/references/framework-zod-lifecycle-and-chainable-api.md +0 -26
  196. package/dist/knowledge/skills/resource-ui-behavior/references/framework-display-edit-modes-api.md +1 -1
  197. package/dist/knowledge/skills/resource-ui-behavior/references/framework-frontend-resource-ui-contracts-api.md +0 -109
  198. package/dist/knowledge/skills/resource-ui-behavior/references/framework-per-primitive-ui-api.md +1 -1
  199. package/dist/knowledge/skills/spec-companion-exports/references/framework-companion-exports-contract-api.md +0 -36
  200. package/dist/knowledge/topic-index.json +128 -285
  201. package/dist/tsconfig.build.tsbuildinfo +1 -1
  202. package/dist/tsconfig.tsbuildinfo +1 -1
  203. package/package.json +4 -4
  204. package/dist/esm/.builder.pid +0 -9
  205. package/dist/knowledge/playbooks/bottom_up.placement-plan/method.md +0 -185
  206. package/dist/knowledge/playbooks/bottom_up.placement-plan/playbook.json +0 -66
  207. package/dist/knowledge/playbooks/top_down.market-evidence-adequacy/method.md +0 -56
  208. package/dist/knowledge/playbooks/top_down.market-evidence-adequacy/playbook.json +0 -40
  209. package/dist/knowledge/playbooks/top_down.market-research/method.md +0 -68
  210. package/dist/knowledge/playbooks/top_down.market-research/playbook.json +0 -42
  211. package/dist/knowledge/playbooks/top_down.market-snapshot-produce/method.md +0 -239
  212. package/dist/knowledge/playbooks/top_down.market-snapshot-produce/playbook.json +0 -69
  213. package/dist/knowledge/references/placement-plan.config.md +0 -330
  214. package/dist/knowledge/references/saas.frontend.settings-hub.engine-destinations.md +0 -58
  215. package/dist/knowledge/references/saas.frontend.settings-hub.placement.md +0 -118
  216. package/dist/knowledge/references/zod.decorators.data-categories.md +0 -197
  217. package/dist/knowledge/rules/resource-schema-awareness.md +0 -77
  218. package/dist/knowledge/skills/resource-config-wiring/references/framework-personal-data-declaration-api.md +0 -622
  219. package/dist/knowledge/skills/resource-schema-authoring/references/framework-personal-data-category-api.md +0 -245
  220. package/dist/knowledge/skills/settings-placement-authoring/SKILL.md +0 -134
  221. package/dist/knowledge/skills/settings-placement-authoring/references/framework-settings-destinations-catalog.md +0 -65
  222. package/dist/knowledge/skills/settings-placement-authoring/references/framework-settings-placement-api.md +0 -127
  223. /package/dist/knowledge/playbooks/{top_down.market-snapshot-produce → top_down.market-analysis}/structured-artifact-action.json +0 -0
@@ -1,737 +1,17 @@
1
1
  # Compliance document candidate method
2
2
 
3
- You are drafting ONE governed document for this application, from what the application has already
4
- declared. Follow the `compliance-document-authoring` skill for the rules; this method says which
5
- document, what shape, and what the two closed vocabularies are.
6
-
7
- ## Choose the document, from a closed set
8
-
9
- `documentRef` must be one of six, and the contract refuses anything else:
10
-
11
- | `documentRef` | what it is |
12
- |---|---|
13
- | `document:public.privacy-notice` | what personal data the application holds and what happens to it |
14
- | `document:public.service-terms` | the commercial and service relationship |
15
- | `document:governance.scope-system-description` | what the system IS, for an assurance reader |
16
- | `document:governance.roles-responsibilities` | who is accountable for what |
17
- | `document:governance.program-handbook` | how the governance program runs |
18
- | `document:governance.document-record-control` | how governed documents are themselves controlled |
19
-
20
- Draft the **privacy notice** unless the run tells you otherwise. It is the one every application
21
- holding personal data needs, and the only one whose facts the current derivation supplies.
22
-
23
- Naming a document outside this set is not a smaller decision than it looks. Each of the six has a
24
- concern composition the framework measures a draft against, so a seventh name is a document nobody can
25
- state the gaps in — and stating the gaps is the whole point.
26
-
27
- ## One section per concern, and SAY which concern
28
-
29
- Every section carries `concernRef`, from the closed compliance-concern vocabulary. **The composition
30
- for each document is inlined at the end of this method** — one entry per concern, saying what it
31
- asserts, what in the application it derives from, what state it usually reaches, and what closes it
32
- when it cannot. Draft one section per concern, in the order given, and produce **every** one:
33
- a concern nothing can answer becomes a section with a marked clause, never a missing heading.
34
-
35
- Those entries are the difference between a draft that states its gaps and one that guesses at them.
36
- Where a concern's entry names an issue number as the remedy, say so in `incompleteness.closedBy` —
37
- `#89`, `#91` — because a numbered issue is a remedy a reader can open, and "application configuration
38
- owner" is not.
39
-
40
- Use `concernRef: null` only for a section that genuinely answers no concern: a preamble, a contact
41
- block. It is not a way past a concern you found hard. A concern you cannot establish gets a SECTION,
42
- with a clause marked incomplete — which is the mechanism this whole document exists around.
43
-
44
- The value must be the concern's exact name. A near miss (`retention` for `retention-and-deletion`) is
45
- worse than a null: it reports the concern as unaddressed while the document addresses it.
46
-
47
- ## What a clause may say
48
-
49
- The rule is ASSERT versus MARK, and the skill states it in full. In short: a clause may say "this is
50
- not yet established"; it may not say something untrue.
51
-
52
- - `DERIVED` — established by a fact this application has declared. Cite it in `factRefs`. A `DERIVED`
53
- clause citing a fact the facts file does not declare is REFUSED before the write.
54
-
55
- **A fact the application declared NOT APPLICABLE is a fact, and a clause may rest on it.** *"No
56
- external analytics provider is configured"* is `DERIVED`, not incomplete: the application settled the
57
- question and recorded the answer. Marking such a concern incomplete says the application does not
58
- know something it does know, which is as wrong as overstating. What is refused is citing a fact the
59
- facts file does not declare AT ALL.
60
- - `NEEDS_FRAMEWORK` / `NEEDS_DECISION` / `NEEDS_CONFIGURATION` — mark it, and say in `incompleteness`
61
- what is missing and who closes it. Four distinct suppliers, four distinct remedies; do not merge them.
62
-
63
- **`NEEDS_DECISION` and `NEEDS_FRAMEWORK` are the pair most easily confused, and the difference is
64
- the REMEDY.** `NEEDS_DECISION` means a surface exists and nobody filled it in — the reader goes and
65
- answers. `NEEDS_FRAMEWORK` means there is NOWHERE to record the answer; telling a reader to decide
66
- is telling them to write into a void.
67
-
68
- Measured on 2026-08-30, a produced notice marked operator identity `NEEDS_DECISION` when no surface
69
- for it existed at all, beside retention periods marked identically and genuinely just unfilled. Two
70
- remedies, rendered the same, and a reader could act on neither.
71
-
72
- **So every incomplete clause must NAME THE SURFACE its answer goes to**, or say plainly that none
73
- exists. A gap with a named surface is a task; a gap without one is a request for work nobody has
74
- scoped. The fragment you are working from names the surface where there is one — carry it into
75
- `incompleteness` rather than leaving the reader to find it.
76
-
77
- **`closedBy` must name WHO can reach that surface, and an ENGINE resource is not the application
78
- author's to edit.** `retentionPolicy`, `systemAccessPolicy.exportSubject` and `.dataCategories()`
79
- on an engine resource live in engine source. Writing "application authors must declare
80
- `retentionPolicy` for each affected engine resource" hands the reader a task they cannot perform —
81
- and a produced notice said exactly that, for 83 resources, in the same confident register as the
82
- remedies that ARE actionable. An unreachable remedy is worse than an unnamed one: it looks like
83
- work, gets scheduled, and fails when someone tries it.
84
-
85
- Where the closer is the FRAMEWORK, say so plainly — "the framework must declare these on its own
86
- resources" — and keep it separate from what this application owes.
87
-
88
- Never fill an unresolved fact, infer a legal decision, or describe the draft as approved, compliant,
89
- legally valid, released or published. Locale and jurisdiction are review inputs; leaving them
90
- unresolved is a legitimate answer and never a reason not to draft.
91
-
92
- ## The refs you write, and their exact shape
93
-
94
- Every `…Ref` field is a **namespaced ref**: `namespace:slug`, both halves lowercase, words joined by
95
- hyphens. `clause:retention-duration-not-declared` is one; `clause:Retention Duration` is not, and a
96
- whole draft is rejected for a single malformed one. Ref ARRAYS are arrays of those strings — never
97
- objects, never `{ ref: … }`.
98
-
99
- Per clause, all three ref arrays are required, and two need at least one entry:
100
-
101
- - `factRefs` — the facts this clause rests on, exactly as the facts file declares them. It may be
102
- EMPTY for a clause that establishes nothing (a marked clause often does).
103
- - `sourcePolicyRefs` — at least one: what governs the statement. Use the fact authority the clause
104
- rests on where there is one (`document-fact-authority:<slug>` from the facts file), otherwise the
105
- policy area it belongs to (`policy:retention`, `policy:recipients`).
106
- - `claimGuardRefs` — at least one: what this clause must NOT be read as claiming. These are the
107
- boundaries you are already holding — `claim-guard:no-legal-applicability-claim`,
108
- `claim-guard:no-invented-retention-period`, `claim-guard:known-fact-only`. Name the ones that apply
109
- to this clause; do not repeat the same guard on every clause out of habit.
110
-
111
- `reviewChecklistRefs` needs at least one (`review-checklist:<slug>`) — what a human must confirm before
112
- this draft is worth anything. `limitationRefs` may be empty, but rarely should be: it is where you say
113
- what the document does NOT cover, in your own words as the drafter, rather than leaving the reader to
114
- discover it.
115
-
116
- ## Where the draft goes
117
-
118
- It lands as a FILE in the application's specifications package, as an ordinary working edit. It is
119
- accepted when a human commits it, and by nothing else. Landing is not approval, release or publication.
120
-
121
-
122
- ---
123
-
124
- # The compositions
125
-
126
- Each of the following is one document's full concern list. Use the one matching your `documentRef`
127
- and ignore the rest.
128
-
129
- ## `document:public.privacy-notice` — thirteen concerns
130
-
131
- ### Concern: operator identity
132
-
133
- **Asserts** — who legally operates the service: entity name, registered address, contact route for
134
- data-protection questions, and a representative where one is required.
135
-
136
- **Derives from** — `operatorComplianceIdentity` in `wildo.saas.config.ts`: the legal entity, its
137
- registered address, the privacy contact route, and an Art. 27 representative where one is required.
138
-
139
- The VALUE is never derivable — a legal entity is a fact about the world, not a property of a codebase.
140
- What the framework provides is the PLACE to record it, and until 2026-08-30 it did not: this fragment
141
- said "no engine change would close it", which was right about derivation and wrong about the home.
142
- The consequence was concrete. A landing REPLACES the produced document on every run, so an operator
143
- who answered by editing the document lost the answer at the next regeneration — the answer has to live
144
- in the application or it does not survive.
145
-
146
- **Completeness** — `DERIVED` when the identity is declared; `NEEDS_DECISION` when it is not.
147
-
148
- ⚠️ `NEEDS_DECISION`, not `NEEDS_FRAMEWORK`, and the distinction is load-bearing now that the surface
149
- exists. `NEEDS_FRAMEWORK` means there is nowhere to record the answer; `NEEDS_DECISION` means there is
150
- and nobody has. They have different remedies — build something, versus fill something in — and a
151
- reader who cannot tell them apart cannot act on either. Say which surface the answer goes to.
152
-
153
- **What closes it** — the operator declaring `operatorComplianceIdentity`. It is cheap and it can be
154
- filled in later without redrafting anything around it, which is the reason it is not a blocker.
155
-
156
- **Never infer it.** Not from the application name, not from a domain, not from another application,
157
- not from model memory. A published legal document naming the wrong entity is worse than one that says
158
- the entity is not yet recorded — the first misdirects a person exercising a right, the second tells
159
- them the document is incomplete.
160
-
161
- **Every document needs this fragment**, which makes it the clearest demonstration of why producing
162
- with gaps beats blocking: the whole document set would be unproducible until one form is filled in.
163
-
164
-
165
- ### Concern: legal basis and controller/processor role
166
-
167
- **Asserts** — the lawful basis for each processing purpose, and whether the operator acts as
168
- controller or processor for it.
169
-
170
- **Derives from** — `operatorComplianceIdentity.lawfulBases` and `.roleAllocations` in
171
- `wildo.saas.config.ts`.
172
-
173
- The DETERMINATIONS are never derivable — both are legal conclusions about the operator's own
174
- arrangements, and the framework can show WHAT is processed and BY WHICH MECHANISM without establishing
175
- under which basis or in which role. What it now provides is where to record the conclusion, so that a
176
- regeneration reproduces it instead of destroying it.
177
-
178
- **The role is a LIST, not a field, and the produced notice said so before the surface existed:**
179
- *"This notice therefore does not assign a single legal role to all data processed in a workspace."* A
180
- B2B SaaS is typically a processor for the records its customers put in and a controller for its
181
- account holders' own data, simultaneously. `roleAllocations` carries one role per population, and a
182
- document that collapses them into one is wrong about both.
183
-
184
- **`LEGITIMATE_INTERESTS` cannot be declared without its assessment** — the schema refuses it. It is
185
- the only Art. 6(1) basis whose lawfulness depends on a balancing test, so the bare claim asserts a
186
- conclusion nobody reached.
187
-
188
- **Completeness** — `DERIVED` when declared; `NEEDS_DECISION` when not. Never `NEEDS_FRAMEWORK`: there
189
- is a surface, so the remedy is to fill it in, not to build something. Name the surface when reporting
190
- the gap.
191
-
192
- **What closes it** — the operator determining, per purpose, the basis and the role, ideally with
193
- counsel, and declaring them in `operatorComplianceIdentity`. The tenancy fragment describes the SHAPE this usually takes in a B2B SaaS — processor for
194
- the data a customer stores about its own people, controller for the data the operator holds about
195
- account holders — but describing the common shape is not the same as determining this operator's, and
196
- the produced document must not present the first as the second.
197
-
198
- **This is the fragment most likely to be filled in by a confident guess**, because the shape is
199
- familiar and the sentence is easy to write. Resist it. A basis stated in a published notice is a
200
- position the operator has taken; a basis that was never determined is a position nobody can defend.
201
-
202
-
203
- ### Concern: personal data held
204
-
205
- **Asserts** — which of the application's own records are about identifiable people, and which hold no
206
- personal data of their own.
207
-
208
- **Derives from** — each resource's `retentionPolicy.mode` (`ErasureRetentionMode`, in
209
- `@wildo-ai/saas-models`), read for every resource the application declares:
210
-
211
- | declaration | what the document may state |
212
- |---|---|
213
- | `RETAIN_AND_IMPERSONALIZE` | this record holds personal data; the fields marked `.impersonalizeWith()` are what is scrubbed on erasure |
214
- | `RETAIN_ONLY` | this record holds **no personal data of its own** — a positive, factory-enforced statement, because declaring a scrub treatment under this mode fails at config build |
215
- | absent | the application has not said |
216
-
217
- Read `dataSubjectPolicy.kind` for the SESSION question only. `NO_PRINCIPAL` is equally true of an
218
- invoice and of a CRM contact, so it never tells you whether a record is about a person — the
219
- retention mode does.
220
-
221
- **Completeness** — `DERIVED` for every resource that declares a retention policy.
222
-
223
- **An erasure-coverage gap does NOT license a claim about whether personal data is held.** The two are
224
- different axes and the facts report them separately: `retentionPolicy` answers *what happens to this
225
- row on erasure*, while a field's `.dataCategories()` answers *what kind of data is in it*. A resource
226
- can declare the second and not the first, and 48 of the engine's 85 do.
227
-
228
- So writing "N resources have no declared erasure treatment, so this notice cannot establish whether
229
- they hold personal data" is an over-reach, and it contradicts the data-categories section of the same
230
- document — which enumerates categories drawn from those very resources. It was produced twice on
231
- 2026-08-30 before this paragraph existed.
232
-
233
- State the gap as what it is: those resources have not declared an erasure TREATMENT, so a notice
234
- cannot say what an erasure does to them. Where the facts also report a category coverage for the same
235
- origin, say what IS established alongside it.
236
- `NEEDS_CONFIGURATION` for every resource that does not.
237
-
238
- **What closes it** — the application's own authors declaring `retentionPolicy` on the resource. Name
239
- the resources individually: the remedy is per-resource, and a count is not actionable.
240
-
241
-
242
- ### Concern: data categories
243
-
244
- **Asserts** — which CATEGORIES of personal data are held: identity, contact details, credentials,
245
- financial, usage, location, user content, professional, and any GDPR Art. 9 special-category or
246
- Art. 10 criminal-offence data.
247
-
248
- **Derives from** — each field's `.dataCategories()` declaration, reported per resource and unioned
249
- across the application in the compliance-primary-facts section. One category is DERIVED rather than
250
- declared: `credentials`, from `.isPassword()` / `.isOtp()` / `.isWriteOnlySecret()`, which are
251
- markers the author already wrote.
252
-
253
- **Nothing is derived from a field NAME.** `email`, `phone` and `dateOfBirth` are suggestive while
254
- `notes` is not, so a category list assembled by pattern-matching identifiers is a guess presented as
255
- an inventory — and its failure mode is silent, because whatever the guess misses is simply not
256
- disclosed. The declaration exists so that never happens.
257
-
258
- **Completeness** — `DERIVED` when the application declares categories; `NEEDS_CONFIGURATION` when it
259
- declares none, or when the facts section lists OUTSTANDING fields.
260
-
261
- **The two gaps are different and must not be written the same way.**
262
-
263
- - A resource that declares no category at all may genuinely hold no personal data. Nothing
264
- distinguishes "considered and not applicable" from "never asked", so state that the question is
265
- unanswered — never that the application holds no personal data.
266
- - A field the facts section lists as OUTSTANDING is a provable omission. The application already
267
- declared that field personal by another door — an erasure scrub treatment, a portability
268
- provenance, or a credential marker — so the missing category is an omission, not a judgement.
269
- Name the field and its basis. Do not guess what it contains.
270
-
271
- **Say WHICH special category, not that there is one.** A `special-category:*` value is Art. 9 data
272
- and carries a lawful-basis condition of its own beyond Art. 6; `criminal-offence` is Art. 10, a
273
- separate article with a different condition. "We process special category data" is not a disclosure —
274
- the enum names the kind precisely so a notice can.
275
-
276
- **Do NOT point at `Compliance_DataClassification`.** It exists, it is a four-level SENSITIVITY ladder
277
- (public / internal / confidential / restricted) applied only to flows-actors configuration, and it
278
- answers "how protected must this be" rather than "what kind of personal data is this". Using it here
279
- would produce a notice describing protection levels where a reader expects data types. It is named
280
- explicitly because it is a near-miss a reader will otherwise find and assume answers this concern.
281
-
282
-
283
- ### Concern: consent and preferences
284
-
285
- **Asserts** — what the application asks consent for, at what granularity, and what a person's choice
286
- actually controls.
287
-
288
- **Derives from**
289
-
290
- - `DataProcessingConsent` (`saas-models` → `compliance/privacy/privacy.shared.schemas.ts`):
291
- `NONE` · `BASIC` · `ANALYTICS` · `MARKETING` · `FULL` — a ladder, so a level implies the ones below it.
292
- - `analytics.requiredConsent` (`app-configuration-shared.shared.schemas.ts`, default `ANALYTICS`) —
293
- the minimum consent level for ANALYTICS, and nothing wider.
294
- - per-user `dataProcessingConsent` — the level each person actually chose.
295
-
296
- ⚠️ **`requiredConsent` is ANALYTICS-SCOPED, and this fragment claimed otherwise until 2026-08-30.**
297
- It read "the minimum a person must give to use the service", which is a different and much larger
298
- claim; the field sits inside `AnalyticsConfigSchema` and governs analytics alone. A drafter following
299
- the old wording would have published that a person must consent to analytics in order to use the
300
- product — false, and false about the one subject where a reader is most likely to act on it.
301
-
302
- - `operatorComplianceIdentity.serviceConsentRequirement` (`wildo.saas.config.ts`) — what a person
303
- must agree to in order to use the service AT ALL, and what that gate covers.
304
-
305
- **`NONE` is a POSITIVE answer and the usual one.** The operator declaring that nothing is gated on
306
- consent has answered the question; a document must state it as settled rather than treating it as an
307
- absence. ABSENT is the different case — nobody was asked — and the brief distinguishes the two.
308
-
309
- **A consent GATE is not a lawful BASIS.** For most B2B SaaS the basis for core processing is CONTRACT:
310
- a person uses the product because their employer signed for it. A notice that reads a gate as a basis
311
- states that their data is processed on consent they could withdraw, which is wrong about the single
312
- most consequential thing in the document. The basis is declared per purpose in `lawfulBases`.
313
-
314
- **Completeness** — `DERIVED` for the analytics gate, the per-user level, and the service-wide
315
- requirement once declared; `NEEDS_DECISION` naming `serviceConsentRequirement` when it is not.
316
-
317
- This was `NEEDS_FRAMEWORK` until 2026-08-30, correctly: there was nowhere to record the answer. Issue
318
- #139 built the surface, so the remedy changed from "someone must build this" to "fill this in" — a
319
- different party and a different task, which is why the two states are never interchangeable.
320
-
321
- **Say the thing readers get wrong.** Consent is one lawful basis among several, and it is not the one
322
- most B2B processing rests on. **Never write that all processing rests on consent, and never write
323
- that accepting terms supplies a lawful basis** — accepting a contract is a different basis with
324
- different consequences, and conflating them is the single most common defect in a generated privacy
325
- notice. Which basis applies is a `NEEDS_DECISION`; see the legal-basis fragment.
326
-
327
-
328
- ### Concern: retention and deletion
329
-
330
- **Asserts** — what happens to a record when its subject is erased, and how long records are kept.
331
-
332
- Two questions, and the framework answers only the first.
333
-
334
- **Derives from (the HOW — answerable)** — each resource's `retentionPolicy.mode`:
335
- `RETAIN_AND_IMPERSONALIZE` keeps the row, freezes it, hides it from normal reads and scrubs the fields
336
- marked `.impersonalizeWith()`; `RETAIN_ONLY` keeps and freezes a row that holds no personal data of
337
- its own. Reading the per-field treatments tells the document exactly WHICH values are destroyed and
338
- which survive, which is a genuinely strong statement most notices cannot make.
339
-
340
- **Derives from (the HOW LONG)** — the application's `dataRetention` schedule, declared in its config
341
- over four data classes: `RESOURCE_DATA` (its records), `FILES` (uploaded bytes), `AUDIT_LOGS` (the
342
- tamper-evident trail, kept longest and for a different reason — it is the evidence an operator may
343
- lawfully retain under GDPR Art. 17(3)) and `OPERATIONAL_LOGS`. Each declared class carries a period in
344
- days AND the operator's stated reason for it.
345
-
346
- The classes are the STORES on purpose: the same declaration answers what a notice must state and what
347
- a backup-and-restore inventory must protect (issue #8).
348
-
349
- **Completeness** — `DERIVED` for the treatment, and for each retention class the application HAS
350
- declared. `NEEDS_DECISION` for each class it has not: the mechanism exists, so what is missing is an
351
- operator's decision rather than an engine change. Do not let the derivable half hide the missing one —
352
- a statement about erasure mechanics reads to a reader as an answer about retention, so both go in the
353
- same breath.
354
-
355
- **What closes it** — the operator deciding a period, and saying why. Not a framework change: issue #89
356
- delivered the declaration surface, so an undeclared class is now an unanswered question rather than an
357
- unaskable one.
358
-
359
- **An UNDECLARED class is not "kept indefinitely" and not "deleted immediately".** It is a decision
360
- nobody has taken, and the document must say exactly that.
361
-
362
- **Never write a placeholder period.** "Typically 12 months" in a published notice is a commitment
363
- nobody made and nothing enforces.
364
-
365
-
366
- ### Concern: recipients and processors
367
-
368
- **Asserts** — who else receives personal data, in what role, and where they process it.
369
-
370
- **Derives from (partial)** — the engine capabilities the application has a provider configured for
371
- (`EngineCapability` in `saas-models` → `external-providers/engine-capabilities.shared.schemas.ts`).
372
- That establishes THAT a third party is involved and WHAT FUNCTION it performs — email delivery,
373
- storage, payments, LLM inference, observability.
374
-
375
- **Derives from (the identities)** — the application's PROCESSOR REGISTER, which your brief renders
376
- under *"Who else processes personal data"*. Per entry: the processor's **legal entity**, the
377
- **regions** it processes in, and its own **subprocessors** with their regions.
378
-
379
- **The register holds two kinds of entry and the brief labels them differently — say which.** A
380
- CONFIGURED PROVIDER entry joins to the application's composition, so it can be checked against it: a
381
- claim about a provider the application does not configure is detectable, and a configured provider
382
- with no entry is reported as UNDECLARED. An AUTHORED entry — a hosting company, a CDN, a support desk
383
- — joins to nothing, because no configuration mentions it. The framework can neither confirm it nor
384
- notice its absence.
385
-
386
- Both are real processors and both belong in the notice. What must not happen is presenting the second
387
- with the authority of the first: an authored entry is the operator's word, and a document that reads
388
- as though the framework verified it overstates what anyone checked.
389
-
390
- **A SELF-HOSTED provider is not a recipient either, and this is the second way to get the section
391
- wrong.** The brief's *"Who RUNS each bound provider"* block splits runtime-bound providers three ways:
392
- THIRD-PARTY RECIPIENTS (vendor-hosted, so data reaches a third party — enumerate these with their
393
- region), OPERATOR-RUN (self-hosted, so no third party receives anything — state them as processing
394
- under the operator's own control, never as recipients), and UNANSWERED.
395
-
396
- An UNANSWERED provider is bound to a runtime with no declared deployment model. **Do not assume
397
- vendor-hosted.** It looks like the cautious reading and it is not: it publishes that a person's data
398
- goes to a third party when nobody said so. Report it as not established, and name the surface —
399
- `providerComplianceChoices` in `wildo.saas.config.ts`.
400
-
401
- **A provider PROFILE is not a recipient.** The brief separates providers BOUND TO A RUNTIME — backend,
402
- companion, workers, minions, a frontend service, which can process the application's data — from
403
- provider profiles with no runtime binding. A profile says how to reach a provider; build-time tooling
404
- lives there, and a coding agent operating on the workspace receives no customer records. Never name a
405
- profile-only provider as a recipient. Naming one reads as caution and is a false statement about
406
- somebody's data, exactly like omitting a real one.
407
-
408
- The section joins the register to the RUNTIME-BOUND providers, so it tells you three different things
409
- and they are not interchangeable:
410
-
411
- - a provider WITH an entry — state its legal entity and regions. `DERIVED`.
412
- - a provider listed **UNDECLARED** — it IS in use and nobody has established who it is. Name the
413
- function it performs (the capability says that) and state that its identity and location are
414
- outstanding. `NEEDS_DECISION`, closed by the operator establishing it.
415
- - a provider listed **STALE** — declared but no longer configured. Do NOT name it as a recipient; it
416
- receives nothing.
417
-
418
- **Completeness** — `DERIVED` per provider that is declared, `NEEDS_DECISION` per provider that is not.
419
- Not `NEEDS_FRAMEWORK` any more: issue #88 delivered the declaration surface, so an undeclared processor
420
- is an unanswered question rather than an unaskable one.
421
-
422
- **Never infer a region from a vendor's brand.** "It is a US company" is not a processing location, and
423
- a transfer statement built on it is a claim about somebody's data made from a guess.
424
-
425
- **Where the register is empty, the honest statement names the FUNCTIONS and admits the identities are
426
- outstanding.**
427
- "Third parties process data for email delivery, file storage and payments; their identities and
428
- processing locations are not yet established in this document" is true, useful and specific. What is
429
- forbidden is the other two moves: inventing a plausible vendor list, and omitting the section — the
430
- first is false and the second lets a reader assume there are no recipients at all.
431
-
432
-
433
- ### Concern: international transfers
434
-
435
- **Asserts** — whether personal data leaves its origin jurisdiction, to where, and under which transfer
436
- mechanism.
437
-
438
- **Derives from** — the processor register's regions, at BOTH levels: each declared processor and
439
- each processor's own subprocessors. Your brief renders them under *"Personal data reaches processors
440
- in: …"*.
441
-
442
- Both levels, because the chain is what makes a transfer statement hard. An operator in France using a
443
- French vendor that runs on a US cloud HAS a transfer, and reading only the vendor's region misses it.
444
-
445
- **The question has three inputs and the framework holds ONE.** Where declared processors process is
446
- derivable. Which jurisdiction is the ORIGIN, and which MECHANISM makes a given transfer lawful, are
447
- legal determinations with a shelf life — adequacy decisions move by court ruling, so a list baked into
448
- the engine would go stale silently and emit a false compliance statement. That is the worst failure
449
- this lane has, so it is deliberately not modelled.
450
-
451
- **Completeness** — three states, and they are three different claims. Do not round any of them up.
452
-
453
- | the register | what you may state | completeness |
454
- |---|---|---|
455
- | ABSENT | nothing — where data lands is unknown | `NEEDS_CONFIGURATION` |
456
- | declares NO processor | no personal data reaches a third party, so none is transferred THROUGH one | `DERIVED`, with the residual stated |
457
- | declares processors | where data lands, region by region | `DERIVED` for location; the transfer POSITION stays `NEEDS_DECISION` |
458
-
459
- **The residual is not optional wording.** "No processor receives personal data" establishes nothing
460
- about the operator's OWN processing — its infrastructure, its staff, its own establishment. A document
461
- that says "we do not transfer data internationally" on the strength of an empty register has answered
462
- a larger question than it was asked.
463
-
464
- **What closes the rest** — the transfer MECHANISM (adequacy, standard contractual clauses, binding
465
- corporate rules, a derogation), recorded per subprocessor entry by a human informed by research. It is
466
- chosen and documented rather than derived, and it is parked deliberately: transfers arrive late in a
467
- company's life, and building the model before a real case exists would be guessing at its shape.
468
-
469
- **"We do not transfer data internationally" is a claim, not a default.** It may be true, and it is
470
- citable when an application has declared it as a fact — but it must never be produced because nothing
471
- said otherwise. Absence of a region declaration means the region is unknown, which is the opposite of
472
- knowing it stays put.
473
-
474
-
475
- ### Concern: automated decisions and AI
476
-
477
- **Asserts** — where the application uses AI, on what data, and whether any of it decides something
478
- about a person.
479
-
480
- **Derives from (partial)** — the AI capabilities the application has a provider configured for:
481
- `AI_LLM`, `AI_EMBEDDINGS`, `AI_DOCUMENT_EXTRACTION`, `AI_CODING_AGENT`. That establishes that AI is in
482
- use and for what function.
483
-
484
- **Not derivable** — whether any of it produces a **decision with a legal or similarly significant
485
- effect** on a person, and whether a human reviews it. Nothing in the framework declares consequence or
486
- human-review posture, and neither can be inferred from a capability being enabled.
487
-
488
- **Derives from (the consequence)** — the AI POSTURE section of your brief, per ENABLED AI capability:
489
- what it is used for, whether it `NO_DECISION` / `ASSISTS_HUMAN_DECISION` / `DECIDES_AUTONOMOUSLY`, and —
490
- where it decides anything — how a person obtains human intervention.
491
-
492
- An application enabling NO AI capability gets a plain statement to that effect: it performs no AI
493
- processing of personal data. That is a positive, checkable answer, not a gap.
494
-
495
- **Completeness** — `DERIVED` per capability with a declared posture; `NEEDS_DECISION` per capability
496
- listed OUTSTANDING. Not `NEEDS_FRAMEWORK` any more: issue #92 delivered the declaration, so an
497
- undeclared posture is an unanswered question rather than an unaskable one.
498
-
499
- **An OUTSTANDING capability is not a harmless one.** This is the one concern where silence is more
500
- dangerous than elsewhere: a notice describing AI use and saying nothing about consequence does not read
501
- as "unknown", it reads as "nothing significant happens here" — which is the exemption, claimed for
502
- free. State the gap explicitly.
503
-
504
- **Why the distinction carries weight.** Automated decision-making with significant effects attracts
505
- specific obligations — information about the logic, and a right to human intervention. A generated
506
- notice must not silently claim the exemption by describing AI use without addressing consequence. The
507
- honest statement says AI is used for these functions, and that whether any of it decides something
508
- about a person is not yet established.
509
-
510
-
511
- ### Concern: analytics and tracking
512
-
513
- **Asserts** — what is observed about how people use the service, and what a person's consent controls.
514
-
515
- **Derives from (partial)** — the observability and analytics capabilities the application has
516
- configured (`FRONTEND_PRODUCT_ANALYTICS`, `LLM_OBSERVABILITY`, `OBSERVABILITY_TELEMETRY`), together
517
- with the consent ladder: `requiredConsent` and per-user `dataProcessingConsent` make
518
- `DataProcessingConsent.ANALYTICS` a real, enforced boundary rather than a stated intention.
519
-
520
- **Derives from (the inventory)** — the CLIENT STORAGE section of your brief: each cookie and storage
521
- key, its mechanism, purpose, necessity and lifetime, split by who sets it.
522
-
523
- The FRAMEWORK's own entries are listed first and belong in every application's table — `__wildo_refresh`
524
- and `__wildo_anon_session`, both strictly necessary, both with lifetimes taken from the application's
525
- session configuration. **No application authored these and every application sets them**, so a table
526
- built only from what the application declared would omit the session cookie, which is the entry a
527
- visitor is most certain to have.
528
-
529
- Then, per frontend surface: its declared entries, an explicit "sets nothing of its own", or
530
- UNINVENTORIED.
531
-
532
- **UNINVENTORIED is not "sets nothing".** The first says nobody catalogued that surface; the second is a
533
- positive statement. Writing the second from the first tells a visitor they are tracked by less than they
534
- are, which is the under-reporting direction a notice must never fail in.
535
-
536
- **Completeness** — `DERIVED` for the capability and consent posture, and for each surface that IS
537
- inventoried; `NEEDS_DECISION` per uninventoried surface. Not `NEEDS_FRAMEWORK` any more: issue #91
538
- delivered the declaration surface, so an uninventoried surface is an unanswered question rather than an
539
- unaskable one.
540
-
541
- Say both halves: which analytics are in use and what consent gates them (strong, derived, enforced), and
542
- what the table still lacks — naming the surfaces, not the concern.
543
-
544
-
545
- ### Concern: subject rights
546
-
547
- **Asserts** — which rights the application can actually honour, and through which mechanism. Not what
548
- the law grants; what this application will do when asked.
549
-
550
- **Derives from**
551
-
552
- - `SubjectExportPurpose.ACCESS_ART_15` and `PORTABILITY_ART_20`
553
- (`saas-models` → `compliance/privacy/subject-export.shared.schemas.ts`) — the two distinct export
554
- bundles. Access is the broader one: every row concerning the subject, including rows they do not
555
- own, such as an audit entry about them or a membership an administrator created.
556
- - each resource's `systemAccessPolicy.exportSubject` — whether an export can answer for that resource
557
- at all. **Absent means the resource is excluded from every export**, so an application that has not
558
- declared it cannot honour Art. 15 over that data.
559
- - `portabilityPolicy` and per-field `.portabilityProvenance()` — Art. 20 covers what the subject
560
- supplied, not what the controller derived about them, and an Art. 20 export refuses on any field
561
- that has declared neither.
562
- - erasure: `eraseSubjectAsSystem(resourceType, rowIds)`, which takes a resource type and row ids and
563
- needs no session. **An account is not required to be a data subject** — a contact, a lead or a
564
- subscriber is erased by exactly this path.
565
-
566
- **Completeness** — `DERIVED` where the resources hold `exportSubject` declarations.
567
-
568
- **What closes it** — per-resource `exportSubject` and, for portability, a `portabilityPolicy` default
569
- plus provenance on the fields that differ from it.
570
-
571
- **Never state** that a right is honoured for a resource whose `exportSubject` is absent. The export
572
- will not include it, and the notice would be describing a capability the application does not have.
573
-
574
-
575
- ### Concern: data exit
576
-
577
- **Asserts** — what a customer gets back when they leave, in what form, and what happens to what
578
- remains.
579
-
580
- **Derives from** — `SubjectExportPurpose.PORTABILITY_ART_20` and the resources declaring
581
- `exportSubject`; the account-closure path; and each resource's `retentionPolicy.mode`, which decides
582
- whether a record is removed or **retained and impersonalized** on erasure.
583
-
584
- **Completeness** — `DERIVED` for the mechanism.
585
-
586
- **Retained-and-impersonalized is the part a reader must not be surprised by**, and the part a
587
- generated document most often gets wrong by omission. A retained row is kept, frozen, hidden from
588
- normal reads and reachable only through an audited privileged path — it is not deleted. Say so. A
589
- notice that implies everything is erased, when financial and audit records are lawfully retained with
590
- their personal fields scrubbed, is inaccurate in the direction that damages trust when discovered.
591
-
592
- **How LONG anything is retained is not derivable** — see the retention-and-deletion fragment.
593
-
594
-
595
- ### Concern: tenancy model
596
-
597
- **Asserts** — that customer data is separated per organization, and where the boundary between the
598
- operator's responsibilities and the customer's lies.
599
-
600
- **Derives from** — the `ORGANIZATIONS` engine capability, the organization scoping applied to every
601
- tenant-scoped resource, and the cross-tenant admission rules (a crossing must be declared on the
602
- operation and carries a usable access grant).
603
-
604
- **Completeness** — `DERIVED`.
605
-
606
- **This fragment is where shared responsibility is described, and it must never be flattened.** In a
607
- B2B SaaS the operator is a PROCESSOR for the data a customer stores about its own people, and a
608
- CONTROLLER for the data it holds about the customer's users as account holders. Those are different
609
- roles with different duties over the same database. Say which is which; do not pick one and apply it
610
- to everything.
611
-
612
- **The attribution has a home now.** `operatorComplianceIdentity.roleAllocations` in
613
- `wildo.saas.config.ts` carries one role per population of data — PROCESSOR for the records a customer
614
- stores about its own people, CONTROLLER for the account data of the people who administer them. It is
615
- a LIST for exactly the reason this fragment gives: those are different roles over the same database,
616
- and a single field would force one of them to be wrong.
617
-
618
- Where the attribution is not established, that is a `NEEDS_DECISION` — never a guess — and the
619
- surface to name is `roleAllocations`. Not `NEEDS_FRAMEWORK`: there is somewhere to write the answer,
620
- so the remedy is to fill it in rather than to build something. See the legal-basis fragment.
621
-
622
-
623
- ## `document:public.service-terms` — six concerns
624
-
625
- ### Concern: service scope
626
-
627
- **Asserts** — what the service actually is: the modules it runs, the resources it manages, and the
628
- surfaces it exposes.
629
-
630
- **Derives from** — the application's declared modules and their resources, and the surfaces it builds
631
- (application, API, public website, technical documentation).
632
-
633
- **Completeness** — `DERIVED`.
634
-
635
- **Derive it; do not describe the product from its marketing.** Scope in a governed document is the
636
- boundary of what the statements around it apply to, so an aspirational description silently widens
637
- every other clause. The declared modules are the honest boundary, and a capability the application
638
- does not run does not belong here even if it is on the roadmap.
639
-
640
-
641
- ### Concern: commercial terms
642
-
643
- **Asserts** — what is sold, how it is charged, and what happens on renewal, change and cancellation.
644
-
645
- **Derives from** — the application's billing declarations
646
- (`saas-models` → `billing/billing-enums.shared.schemas.ts`): `ProductType`, `BillingTiming`,
647
- `ProrationBehavior`, `TaxCalculationMode`, plus the declared products and any usage metering.
648
-
649
- **Completeness** — `DERIVED` where billing is configured; `NEEDS_CONFIGURATION` where the application
650
- declares no products, in which case say that the commercial terms are not yet established rather than
651
- describing a plan nobody has authored.
652
-
653
- **Keep consumer and business audiences apart.** Consumer pre-contract obligations (withdrawal
654
- periods, pre-contractual information) apply only when a declared audience includes consumers. A terms
655
- document that blends both into one ambiguous set is worse than one that states which audience it is
656
- for — so read the declared audience before writing a single clause about cancellation rights.
657
-
658
-
659
- ### Concern: governing law and liability
660
-
661
- **Asserts** — which law governs the agreement, where disputes are heard, and how liability is limited.
662
-
663
- **Derives from** — nothing. These are negotiated commercial and legal positions.
664
-
665
- **Completeness** — `NEEDS_DECISION`.
666
-
667
- **What closes it** — the operator deciding, with counsel. Note that liability limits interact with
668
- the declared audience: consumer-facing terms cannot limit liability as freely as business-to-business
669
- ones, so this fragment cannot be settled before the commercial-terms fragment establishes who the
670
- terms are for.
671
-
672
- **Never produce a liability clause from a template.** A limitation of liability that was copied
673
- rather than decided is either unenforceable or committing the operator to something nobody agreed —
674
- and unlike most gaps in a produced document, this one is not discovered until it matters.
675
-
676
-
677
- ## `document:governance.*` — the governance concerns
678
-
679
- ### Concern: security controls
680
-
681
- **Asserts** — the security measures the application actually runs, at the level of detail a notice
682
- needs and no further.
683
-
684
- **Derives from** — the application's authored authentication configuration (methods enabled, MFA,
685
- password policy, lockout, step-up, session lifetimes), its access floors, its encryption of stored
686
- secrets, and the `ANTIVIRUS_SCANNING` engine capability where a provider is configured for it.
687
-
688
- **Completeness** — `DERIVED`, bounded.
689
-
690
- **Bounded is the operative word.** A public notice states that measures exist and their general
691
- nature. **It must not enumerate the control set**: a published list of exactly which protections are
692
- in place, and by implication which are not, is a map for somebody attacking the service. Detail
693
- belongs in a security addendum shared under agreement, not on a public page — so this fragment reads
694
- very differently in `public.privacy-notice` than in a security schedule, and that difference is
695
- deliberate rather than an omission.
696
-
697
- **Never state a certification.** "Compliant", "certified", "audited" and "ISO/SOC" claims are not
698
- derivable from configuration and are not this document's to make.
699
-
700
-
701
- ### Concern: audit and accountability
702
-
703
- **Asserts** — that security-relevant actions are recorded, what kinds are recorded, and where those
704
- records can be sent.
705
-
706
- **Derives from** — the `COMPLIANCE_AUDIT_TRAIL` engine capability
707
- (`saas-models` → `external-providers/engine-capabilities.shared.schemas.ts`), the auditable-event
708
- catalog the engine emits (authentication, authorization grants, consent decisions, token lifecycle,
709
- configuration and permission changes, cross-tenant administrative access), and per-organization SIEM
710
- routing where a tenant has configured it.
711
-
712
- **Completeness** — `DERIVED`.
713
-
714
- This is one of the fragments where the ENGINE answers for the application. Every generated
715
- application inherits the event catalog and the emitters; a document may state it without the
716
- application having configured anything. Where a tenant has additionally routed events to its own
717
- SIEM, that is the tenant's configuration and belongs in a shared-responsibility statement rather than
718
- in the operator's own notice.
719
-
720
-
721
- ### Concern: access control and roles
722
-
723
- **Asserts** — who inside the operator and inside a customer tenant can reach personal data, and what
724
- constrains them.
725
-
726
- **Derives from** — `CORE_APP_ROLES` and `CORE_ORG_ROLES`
727
- (`saas-models` → `security/authorizations/roles.shared.schemas.ts`) plus the application's own role
728
- table; the administrative-continuity floors (a tenant cannot be left with no usable owner, the
729
- application cannot be left with no super-administrator); and the platform-access-grant mechanism,
730
- under which a platform operator reaching into a customer tenant requires a granted, time-bounded
731
- authorization rather than standing privilege, and every crossing is audited.
732
-
733
- **Completeness** — `DERIVED`.
734
-
735
- Worth stating plainly in a B2B notice, because it is the question a customer's own reviewer asks
736
- first: **operator access to tenant data is not ambient.** It is declared per operation, requires a
737
- usable grant, and is recorded when it happens.
3
+ Use the `compliance-document-authoring` skill and the exact host-supplied
4
+ blueprint and fact projection. Produce only the structured
5
+ `ApplicationAssuranceDocumentCandidateContentDraftV1` proposal requested by the
6
+ host.
7
+
8
+ Preserve section and clause order, exact semantic refs, fact bindings,
9
+ source-policy refs, claim-guard refs, variant selections, review checklists and
10
+ limitations. Do not fill an unresolved fact, infer a legal decision, copy
11
+ restricted text, or describe the candidate as approved, compliant, legally
12
+ valid, released or published.
13
+
14
+ Locale and jurisdiction are candidate review inputs, not proof of legal
15
+ applicability. Leave every unresolved or unsupported item visible. The trusted
16
+ runtime—not this method—validates the proposal, derives candidate status and
17
+ seals exact content and provenance digests.