feat(use-case-controls): relevant als Stufe statt Hard-Filter + Provenance
CI / test-python-backend (push) Successful in 30s
CI / test-python-document-crawler (push) Has been skipped
CI / guardrail-integrity (push) Has been skipped
CI / test-python-dsms-gateway (push) Has been skipped
CI / secret-scan (push) Has been skipped
CI / dep-audit (push) Has been skipped
CI / sbom-scan (push) Has been skipped
CI / build-sha-integrity (push) Successful in 12s
CI / validate-canonical-controls (push) Successful in 12s
CI / loc-budget (push) Successful in 25s
CI / go-lint (push) Has been skipped
CI / detect-changes (push) Successful in 15s
CI / python-lint (push) Has been skipped
CI / nodejs-lint (push) Has been skipped
CI / branch-name (push) Has been skipped
CI / nodejs-build (push) Successful in 3m9s
CI / test-go (push) Has been skipped
CI / iace-gt-coverage (push) Has been skipped

Der harte relevant=true-Filter versteckte ~25% des Korpus (40.926 Atome),
~70% davon echte Pflichten (500er-Validierung). relevant wird zur Stufe:

- Service: tier-Param (core=Default schuetzt Agent/CRA; all=alles inkl. review),
  ORDER BY relevant DESC; pro Control relevant/tier/source_type
  (own_library bei license_rule=3, sonst derived) + source_regulation/article;
  core_count/review_count. Pure Helper tier_label + source_type (+ Tests).
- Route: optionaler tier-Query (default core) — contract-safe (additiv).
- Frontend: Coverage-Drill-down /sdk/coverage/[useCase] — Kern-Pflichten vs.
  "zur fachlichen Pruefung", je mit Herkunfts-Badge; Uebersicht zeigt Delta.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
This commit is contained in:
Benjamin Admin
2026-06-15 20:58:25 +02:00
parent e140477c0b
commit 926dc02a09
7 changed files with 385 additions and 23 deletions
@@ -0,0 +1,165 @@
import Link from 'next/link'
import {
type ControlsResponse,
type ControlItem,
provenanceLabel,
provenanceBadgeClass,
severityBadgeClass,
splitByTier,
} from '../_helpers'
const BACKEND_URL =
process.env.COMPLIANCE_BACKEND_URL || 'http://backend-compliance:8002'
export const dynamic = 'force-dynamic'
async function getControls(useCase: string): Promise<ControlsResponse | null> {
try {
const res = await fetch(
`${BACKEND_URL}/api/compliance/v1/controls/use-cases/${encodeURIComponent(
useCase,
)}/controls?tier=all&limit=200`,
{ cache: 'no-store' },
)
return res.ok ? ((await res.json()) as ControlsResponse) : null
} catch {
return null
}
}
function ControlsTable({ rows }: { rows: ControlItem[] }) {
return (
<div className="overflow-hidden rounded-lg border border-gray-200">
<table className="min-w-full divide-y divide-gray-200 text-sm">
<thead className="bg-gray-50 text-left text-xs uppercase text-gray-500">
<tr>
<th className="px-4 py-2">Prüfaspekt</th>
<th className="px-4 py-2">Sub-Thema</th>
<th className="px-4 py-2">Schwere</th>
<th className="px-4 py-2">Herkunft</th>
</tr>
</thead>
<tbody className="divide-y divide-gray-100 bg-white">
{rows.map((c) => (
<tr key={c.id}>
<td className="px-4 py-2 text-gray-900">{c.title}</td>
<td className="px-4 py-2 text-xs text-gray-500">
{c.sub_topic || '—'}
</td>
<td className="px-4 py-2">
{c.severity ? (
<span
className={`rounded px-2 py-0.5 text-xs font-medium ${severityBadgeClass(
c.severity,
)}`}
>
{c.severity}
</span>
) : null}
</td>
<td className="px-4 py-2">
<span
className={`rounded px-2 py-0.5 text-xs font-medium ${provenanceBadgeClass(
c.source_type,
)}`}
title={c.source_article || undefined}
>
{provenanceLabel(c)}
</span>
</td>
</tr>
))}
</tbody>
</table>
</div>
)
}
export default async function UseCaseControlsPage({
params,
}: {
params: Promise<{ useCase: string }>
}) {
const { useCase } = await params
const data = await getControls(useCase)
if (!data) {
return (
<div className="mx-auto max-w-7xl space-y-4 p-6">
<Link
href="/sdk/coverage"
className="text-sm text-purple-600 hover:underline"
>
Abdeckung
</Link>
<div className="rounded-lg border border-amber-200 bg-amber-50 p-6 text-amber-800">
Keine atom-grain-Daten für {useCase}" gefunden.
</div>
</div>
)
}
const { core, review } = splitByTier(data.controls)
return (
<div className="mx-auto max-w-7xl space-y-8 p-6">
<div className="flex items-center gap-2 text-sm text-gray-500">
<Link href="/sdk/coverage" className="hover:text-purple-600">
Abdeckung
</Link>
<span>/</span>
<span className="text-gray-900">{data.label}</span>
</div>
<header className="space-y-1">
<h1 className="text-2xl font-bold text-gray-900">{data.label}</h1>
<p className="text-sm text-gray-600">
<span className="font-semibold text-gray-900">
{data.core_count.toLocaleString('de-DE')}
</span>{' '}
Kern-Pflichten ·{' '}
<span className="font-semibold text-gray-900">
{data.review_count.toLocaleString('de-DE')}
</span>{' '}
zur fachlichen Prüfung
</p>
</header>
<section className="space-y-2">
<h2 className="text-lg font-semibold text-gray-900">
Kern-Pflichten ({core.length})
</h2>
{core.length ? (
<ControlsTable rows={core} />
) : (
<p className="text-sm text-gray-500">—</p>
)}
</section>
<section className="space-y-2">
<div>
<h2 className="text-lg font-semibold text-gray-900">
Zur fachlichen Prüfung ({review.length})
</h2>
<p className="text-xs text-gray-500">
Breiter gefasste Aspekte — bewusst eher zu viel als zu wenig. Fachlich
prüfen; nicht zutreffende lassen sich (künftig) als unanwendbar
markieren.
</p>
</div>
{review.length ? (
<ControlsTable rows={review} />
) : (
<p className="text-sm text-gray-500">—</p>
)}
</section>
{data.total > data.controls.length ? (
<p className="text-xs text-gray-400">
Angezeigt: erste {data.controls.length.toLocaleString('de-DE')} von{' '}
{data.total.toLocaleString('de-DE')} Sub-Thema-Filter folgt.
</p>
) : null}
</div>
)
}
@@ -3,9 +3,23 @@ import {
licenseTierBadgeClass,
commercialBadgeClass,
groupUseCases,
provenanceLabel,
provenanceBadgeClass,
splitByTier,
severityBadgeClass,
type UseCaseRow,
type ControlItem,
} from './_helpers'
const ctrl = (over: Partial<ControlItem>): ControlItem => ({
id: 'id',
title: 'T',
relevant: true,
tier: 'core',
source_type: 'derived',
...over,
})
const uc = (over: Partial<UseCaseRow>): UseCaseRow => ({
key: 'x',
label: 'X',
@@ -52,4 +66,44 @@ describe('coverage helpers', () => {
])
expect(groups.map((g) => g.group)).toEqual(['document', 'mystery'])
})
it('provenance label: own library vs derived (with document + article)', () => {
expect(provenanceLabel(ctrl({ source_type: 'own_library' }))).toBe(
'Eigene Bibliothek',
)
expect(
provenanceLabel(
ctrl({ source_type: 'derived', source_regulation: 'DSGVO' }),
),
).toBe('Abgeleitet · DSGVO')
expect(
provenanceLabel(
ctrl({
source_type: 'derived',
source_regulation: 'DSGVO',
source_article: 'Art. 30',
}),
),
).toBe('Abgeleitet · DSGVO Art. 30')
// derived but no document known → graceful fallback
expect(provenanceLabel(ctrl({ source_type: 'derived' }))).toBe('Abgeleitet')
})
it('provenance + severity badge classes', () => {
expect(provenanceBadgeClass('own_library')).toContain('amber')
expect(provenanceBadgeClass('derived')).toContain('blue')
expect(severityBadgeClass('critical')).toContain('red')
expect(severityBadgeClass('high')).toContain('orange')
expect(severityBadgeClass(null)).toContain('gray')
})
it('splitByTier separates core (relevant) from review', () => {
const { core, review } = splitByTier([
ctrl({ id: 'a', relevant: true }),
ctrl({ id: 'b', relevant: false, tier: 'review' }),
ctrl({ id: 'c', relevant: true }),
])
expect(core.map((c) => c.id)).toEqual(['a', 'c'])
expect(review.map((c) => c.id)).toEqual(['b'])
})
})
@@ -79,6 +79,82 @@ export function commercialBadgeClass(commercial: string | null): string {
}
}
// --- Controls drill-down (#80 Stufe-Flip + Provenance) ---
export interface ControlItem {
id: string
control_id?: string | null
title: string
objective?: string | null
severity?: string | null
sub_topic?: string | null
canonical_obligation?: string | null
source_regulation?: string | null
source_article?: string | null
relevant: boolean
tier: 'core' | 'review'
source_type: 'derived' | 'own_library'
}
export interface ControlsResponse {
use_case: string
label: string
group: string
granularity: string
tier: string
total: number
core_count: number
review_count: number
limit: number
offset: number
sub_topic: string | null
subtopic_counts: Record<string, number>
controls: ControlItem[]
}
// Provenance line: own library vs derived-from-document (with the document, and
// article when known). The user wants to see WHERE a derived control came from.
export function provenanceLabel(
c: Pick<ControlItem, 'source_type' | 'source_regulation' | 'source_article'>,
): string {
if (c.source_type === 'own_library') return 'Eigene Bibliothek'
const doc = c.source_regulation?.trim()
if (!doc) return 'Abgeleitet'
const art = c.source_article?.trim()
return art ? `Abgeleitet · ${doc} ${art}` : `Abgeleitet · ${doc}`
}
export function provenanceBadgeClass(sourceType: string): string {
return sourceType === 'own_library'
? 'bg-amber-100 text-amber-800'
: 'bg-blue-100 text-blue-800'
}
export function severityBadgeClass(sev: string | null | undefined): string {
switch ((sev || '').toLowerCase()) {
case 'critical':
return 'bg-red-100 text-red-800'
case 'high':
return 'bg-orange-100 text-orange-800'
case 'medium':
return 'bg-yellow-100 text-yellow-800'
default:
return 'bg-gray-100 text-gray-600'
}
}
// Split into the two display tiers: Kern-Pflichten (relevant) and the
// 'zur Prüfung' tier (shown but flagged) — never hidden.
export function splitByTier(controls: ControlItem[]): {
core: ControlItem[]
review: ControlItem[]
} {
const core: ControlItem[] = []
const review: ControlItem[] = []
for (const c of controls) (c.relevant ? core : review).push(c)
return { core, review }
}
export interface UseCaseGroup {
group: string
label: string
+19 -6
View File
@@ -1,3 +1,4 @@
import Link from 'next/link'
import {
type UseCaseRow,
type CorpusOverview,
@@ -47,6 +48,7 @@ export default async function CoveragePage() {
const groups = groupUseCases(useCases)
const totalRelevant = useCases.reduce((s, u) => s + u.atom_relevant, 0)
const totalAtoms = useCases.reduce((s, u) => s + u.atom_total, 0)
const totalReview = totalAtoms - totalRelevant
return (
<div className="mx-auto max-w-7xl space-y-10 p-6">
@@ -62,8 +64,8 @@ export default async function CoveragePage() {
<section className="grid grid-cols-2 gap-3 sm:grid-cols-4">
<Stat label="Use Cases" value={useCases.length} />
<Stat label="Pflichten (relevant)" value={totalRelevant.toLocaleString('de-DE')} />
<Stat label="klassifizierte Atome" value={totalAtoms.toLocaleString('de-DE')} />
<Stat label="Kern-Pflichten" value={totalRelevant.toLocaleString('de-DE')} />
<Stat label="zur Prüfung" value={totalReview.toLocaleString('de-DE')} />
<Stat label="Quell-Dokumente" value={corpus?.totals.documents ?? 0} />
</section>
@@ -104,19 +106,30 @@ export default async function CoveragePage() {
<tr>
<th className="px-4 py-2">Use Case</th>
<th className="px-4 py-2">Key</th>
<th className="px-4 py-2 text-right">relevant</th>
<th className="px-4 py-2 text-right">klassifiziert</th>
<th className="px-4 py-2 text-right">Kern</th>
<th className="px-4 py-2 text-right">zur Prüfung</th>
<th className="px-4 py-2 text-right">gesamt</th>
<th className="px-4 py-2">Quellen</th>
</tr>
</thead>
<tbody className="divide-y divide-gray-100 bg-white">
{g.rows.map((u) => (
<tr key={u.key} className={u.atom_relevant === 0 ? 'text-gray-400' : ''}>
<td className="px-4 py-2 font-medium text-gray-900">{u.label}</td>
<tr key={u.key} className={u.atom_total === 0 ? 'text-gray-400' : ''}>
<td className="px-4 py-2 font-medium">
<Link
href={`/sdk/coverage/${u.key}`}
className="text-purple-700 hover:underline"
>
{u.label}
</Link>
</td>
<td className="px-4 py-2 font-mono text-xs text-gray-500">{u.key}</td>
<td className="px-4 py-2 text-right font-semibold">
{u.atom_relevant.toLocaleString('de-DE')}
</td>
<td className="px-4 py-2 text-right text-amber-700">
{(u.atom_total - u.atom_relevant).toLocaleString('de-DE')}
</td>
<td className="px-4 py-2 text-right text-gray-500">
{u.atom_total.toLocaleString('de-DE')}
</td>
@@ -51,6 +51,12 @@ async def controls_for_use_case(
use_case: str,
primary_only: bool = Query(False, description="master-grain Fallback: nur Primaerzweck"),
sub_topic: Optional[str] = Query(None, description="atom-grain: nur dieses Sub-Thema"),
tier: str = Query(
"core",
pattern="^(core|all)$",
description="atom-grain: 'core'=nur validierte Kern-Pflichten (Default), "
"'all'=alle inkl. 'zur Prüfung'-Stufe",
),
limit: int = Query(50, ge=1, le=200),
offset: int = Query(0, ge=0),
svc: UseCaseControlsService = Depends(get_use_case_controls_service),
@@ -58,4 +64,4 @@ async def controls_for_use_case(
"""Controls for a topic. Atom-grain (Haiku: relevant + sub_topic) wenn vorhanden,
sonst master-grain Seed."""
with translate_domain_errors():
return svc.controls_for_use_case(use_case, primary_only, limit, offset, sub_topic)
return svc.controls_for_use_case(use_case, primary_only, limit, offset, sub_topic, tier)
@@ -43,6 +43,21 @@ def relevance_score(
return round(min(score, 1.0), 3)
def tier_label(relevant: bool) -> str:
"""Soft tier instead of a hard filter: validated obligations are 'core',
the rest are 'review' — shown but flagged for expert curation. The boundary
'concrete vs. generic' is genuinely fuzzy; hiding 'review' dropped ~25% of
the corpus, much of it real (filter validation 2026-06-15)."""
return "core" if relevant else "review"
def source_type(license_rule: Optional[int]) -> str:
"""Provenance: 'own_library' = self-written (license_rule 3, no commercial
source); 'derived' = lifted from a sourced document (license 1/2, the
document is in source_regulation)."""
return "own_library" if license_rule == 3 else "derived"
# Representative member (most severe, then lowest control_id) carries the
# human-readable title/objective — master_controls.canonical_name is only the
# merge token, so we surface a real member control per master.
@@ -76,8 +91,8 @@ _LIST_SQL = text("""
# per-atom relevance + sub-topic. Far more precise + organized than the master
# seed. Preferred whenever the use-case has been processed.
_ATOM_LIST_SQL = text("""
SELECT ac.control_uuid, ac.sub_topic, ac.canonical_obligation,
cc.control_id, cc.title, cc.objective, cc.severity,
SELECT ac.control_uuid, ac.sub_topic, ac.canonical_obligation, ac.relevant,
cc.control_id, cc.title, cc.objective, cc.severity, cc.license_rule,
cpl.source_regulation, cpl.source_article
FROM atom_classification ac
JOIN canonical_controls cc ON cc.id = ac.control_uuid
@@ -86,9 +101,9 @@ _ATOM_LIST_SQL = text("""
FROM control_parent_links cpl
WHERE cpl.control_uuid = ac.control_uuid LIMIT 1
) cpl ON true
WHERE ac.use_case = :uc AND ac.relevant = true
WHERE ac.use_case = :uc AND (:all = true OR ac.relevant = true)
AND (:sub IS NULL OR ac.sub_topic = :sub)
ORDER BY ac.sub_topic NULLS LAST,
ORDER BY ac.relevant DESC, ac.sub_topic NULLS LAST,
CASE cc.severity WHEN 'critical' THEN 0 WHEN 'high' THEN 1
WHEN 'medium' THEN 2 ELSE 3 END, cc.title
LIMIT :lim OFFSET :off
@@ -147,18 +162,24 @@ class UseCaseControlsService:
limit: int = 50,
offset: int = 0,
sub_topic: Optional[str] = None,
tier: str = "core",
) -> dict[str, Any]:
"""Controls for ``use_case``. Prefers the atom-grain Haiku classification
(precise + sub-topic-organized) when present; falls back to the
master-grain seed otherwise."""
master-grain seed otherwise.
``tier`` (atom-grain only): 'core' = validated obligations only (default,
keeps the agent/CRA callers precise); 'all' = everything incl. the
'review' tier (shown, flagged) so the human browse view loses nothing."""
if not is_valid_use_case(use_case):
raise NotFoundError(f"Unknown use_case '{use_case}'")
uc = REGISTRY[use_case]
lim = min(max(int(limit), 1), 200)
off = max(int(offset), 0)
tier = tier if tier in ("core", "all") else "core"
if self._has_atom_grain(use_case):
return self._atom_grain(uc, lim, off, sub_topic)
return self._atom_grain(uc, lim, off, sub_topic, tier)
# --- master-grain fallback (recall seed) ---
count_sql = (
@@ -204,23 +225,29 @@ class UseCaseControlsService:
).scalar() or 0) > 0
def _atom_grain(
self, uc, lim: int, off: int, sub_topic: Optional[str],
self, uc, lim: int, off: int, sub_topic: Optional[str], tier: str = "core",
) -> dict[str, Any]:
total = self.db.execute(text(
"SELECT count(*) FROM atom_classification "
"WHERE use_case = :uc AND relevant = true "
"AND (:sub IS NULL OR sub_topic = :sub)"
), {"uc": uc.key, "sub": sub_topic}).scalar() or 0
all_flag = tier == "all"
counts = self.db.execute(text(
"SELECT count(*) FILTER (WHERE relevant), "
"count(*) FILTER (WHERE NOT relevant) "
"FROM atom_classification "
"WHERE use_case = :uc AND (:sub IS NULL OR sub_topic = :sub)"
), {"uc": uc.key, "sub": sub_topic}).first()
core_count = int((counts[0] if counts else 0) or 0)
review_count = int((counts[1] if counts else 0) or 0)
total = core_count + review_count if all_flag else core_count
facet = {
row[0]: int(row[1])
for row in self.db.execute(text(
"SELECT COALESCE(sub_topic, '(none)'), count(*) "
"FROM atom_classification WHERE use_case = :uc AND relevant = true "
"FROM atom_classification WHERE use_case = :uc "
"AND (:all = true OR relevant = true) "
"GROUP BY 1 ORDER BY 2 DESC"
), {"uc": uc.key}).fetchall()
), {"uc": uc.key, "all": all_flag}).fetchall()
}
rows = self.db.execute(_ATOM_LIST_SQL, {
"uc": uc.key, "sub": sub_topic, "lim": lim, "off": off,
"uc": uc.key, "all": all_flag, "sub": sub_topic, "lim": lim, "off": off,
}).fetchall()
controls = [
{
@@ -233,11 +260,16 @@ class UseCaseControlsService:
"canonical_obligation": r.canonical_obligation,
"source_regulation": r.source_regulation,
"source_article": r.source_article,
"relevant": bool(r.relevant),
"tier": tier_label(r.relevant),
"source_type": source_type(r.license_rule),
}
for r in rows
]
return {
"use_case": uc.key, "label": uc.label, "group": uc.group,
"granularity": "atom", "total": int(total), "limit": lim, "offset": off,
"granularity": "atom", "tier": tier, "total": int(total),
"core_count": core_count, "review_count": review_count,
"limit": lim, "offset": off,
"sub_topic": sub_topic, "subtopic_counts": facet, "controls": controls,
}
@@ -10,6 +10,8 @@ from compliance.domain import NotFoundError
from compliance.services.use_case_controls import (
UseCaseControlsService,
relevance_score,
source_type,
tier_label,
)
_NET_KW = ("firewall", "tls", "port", "segmentation", "network", "header")
@@ -55,3 +57,17 @@ def test_controls_for_unknown_use_case_raises_not_found():
svc = UseCaseControlsService(db=None) # guard runs before any DB access
with pytest.raises(NotFoundError):
svc.controls_for_use_case("does_not_exist")
def test_tier_label_maps_relevance_to_soft_tier():
assert tier_label(True) == "core"
assert tier_label(False) == "review"
def test_source_type_own_library_vs_derived():
# license_rule 3 = self-written framework, no commercial source
assert source_type(3) == "own_library"
# license 1 (public domain/EU/NIST) and 2 (CC-BY) are derived from a document
assert source_type(1) == "derived"
assert source_type(2) == "derived"
assert source_type(None) == "derived"