WCAG 2.2 AA compliance, ARIA patterns, keyboard navigation, screen reader optimization
日本語の概要は準備中です。原文の説明を表示しています。
GDPR compliance - data subject rights, lawful basis, DPIA, privacy by design, breach notification, consent management, cross-border transfers, PII masking
インストールする前に、エージェントに与えられる指示の中身を確認できます。
| Right | Article | SLA | Implementation |
|---|---|---|---|
| Right of Access | Art. 15 | 30 gun | Data export endpoint |
| Right to Rectification | Art. 16 | 30 gun | Profile edit + audit trail |
| Right to Erasure | Art. 17 | 30 gun | Cascading delete + anonymize |
| Right to Restriction | Art. 18 | 30 gun | Processing flag on record |
| Right to Portability | Art. 20 | 30 gun | Machine-readable export (JSON/CSV) |
| Right to Object | Art. 21 | 30 gun | Opt-out mechanism |
| Automated Decision-Making | Art. 22 | 30 gun | Human review override |
interface DSRRequest {
id: string;
type: 'access' | 'rectification' | 'erasure' | 'restriction' | 'portability' | 'objection';
subjectId: string;
verifiedIdentity: boolean;
receivedAt: Date;
deadline: Date; // receivedAt + 30 gun
status: 'received' | 'verified' | 'processing' | 'completed' | 'rejected';
reason?: string;
}
async function handleDSR(request: DSRRequest): Promise<DSRResponse> {
// Step 1: Identity verification ZORUNLU
if (!request.verifiedIdentity) {
return { status: 'rejected', reason: 'Identity not verified' };
}
// Step 2: Check deadline
const daysRemaining = differenceInDays(request.deadline, new Date());
if (daysRemaining <= 5) {
await alertDPO({ type: 'dsr_deadline_approaching', request });
}
// Step 3: Process by type
switch (request.type) {
case 'access':
return await generateDataExport(request.subjectId);
case 'erasure':
return await executeErasure(request.subjectId);
case 'portability':
return await generatePortableExport(request.subjectId, 'json');
case 'rectification':
return await updateSubjectData(request.subjectId, request.corrections);
case 'restriction':
return await restrictProcessing(request.subjectId);
case 'objection':
return await recordObjection(request.subjectId, request.reason);
}
}
async function executeErasure(subjectId: string): Promise<ErasureResult> {
const erasureLog: ErasureStep[] = [];
await db.transaction(async (tx) => {
// 1. Anonymize user record (yasal zorunluluklar haric)
await tx.users.update({
where: { id: subjectId },
data: {
email: `erased-${hash(subjectId)}@deleted.local`,
name: 'Erased User',
phone: null,
address: null,
dateOfBirth: null,
deletedAt: new Date(),
},
});
erasureLog.push({ table: 'users', action: 'anonymized' });
// 2. Delete personal messages
const deletedMessages = await tx.messages.deleteMany({
where: { userId: subjectId },
});
erasureLog.push({ table: 'messages', action: 'deleted', count: deletedMessages.count });
// 3. Delete sessions and tokens
await tx.sessions.deleteMany({ where: { userId: subjectId } });
await tx.refreshTokens.deleteMany({ where: { userId: subjectId } });
erasureLog.push({ table: 'sessions', action: 'deleted' });
// 4. Anonymize audit logs (log kaydi kalir, kisi bilgisi gider)
await tx.auditLogs.updateMany({
where: { actorId: subjectId },
data: { actorId: 'erased', actorEmail: 'erased' },
});
erasureLog.push({ table: 'auditLogs', action: 'anonymized' });
// 5. Notify third-party processors
await notifyProcessors(subjectId, 'erasure');
// 6. Record erasure for compliance
await tx.erasureRecords.create({
data: {
subjectHash: hash(subjectId),
erasedAt: new Date(),
systems: erasureLog,
},
});
});
return { success: true, steps: erasureLog };
}
| Basis | Article | When to Use | Example |
|---|---|---|---|
| Consent | Art. 6(1)(a) | Optional processing, marketing | Newsletter signup |
| Contract | Art. 6(1)(b) | Necessary for service delivery | Order processing |
| Legal obligation | Art. 6(1)(c) | Required by law | Tax records |
| Vital interests | Art. 6(1)(d) | Life-threatening situations | Emergency contact |
| Public interest | Art. 6(1)(e) | Public authority tasks | Government services |
| Legitimate interest | Art. 6(1)(f) | Business need, balanced with rights | Fraud prevention |
## Data Protection Impact Assessment
**Project:** [proje adi]
**Date:** [tarih]
**DPO Review:** [evet/hayir]
### 1. Processing Description
- What data: [veri turleri]
- Why: [amac]
- How: [islem yontemi]
- Who: [erisim kimlerde]
- How long: [saklama suresi]
### 2. Necessity & Proportionality
- Lawful basis: [hukuki dayanak]
- Data minimization: [minimum veri mi?]
- Purpose limitation: [amac sinirli mi?]
- Storage limitation: [saklama suresi uygun mu?]
### 3. Risk Assessment
| Risk | Likelihood | Impact | Severity | Mitigation |
|------|-----------|--------|----------|------------|
| Unauthorized access | [L/M/H] | [L/M/H] | [L/M/H] | [onlem] |
| Data breach | [L/M/H] | [L/M/H] | [L/M/H] | [onlem] |
| Purpose creep | [L/M/H] | [L/M/H] | [L/M/H] | [onlem] |
### 4. Measures
- [ ] Encryption at rest and in transit
- [ ] Access controls (RBAC)
- [ ] Audit logging
- [ ] Data minimization applied
- [ ] Retention policy configured
- [ ] Pseudonymization where possible
### 5. DPO Sign-off
Date: [tarih] | Approved: [evet/hayir]
// YANLIS: Tum veriyi topla
const user = await db.users.findUnique({
where: { id },
// Returns everything including SSN, DOB, etc.
});
// DOGRU: Sadece gerekli alanlari sec
const user = await db.users.findUnique({
where: { id },
select: {
id: true,
name: true,
email: true,
// SSN, DOB, etc. secilmez
},
});
function maskPII(data: Record<string, unknown>): Record<string, unknown> {
const piiFields: Record<string, (val: string) => string> = {
email: (v) => v.replace(/^(.{2}).*(@.*)$/, '$1***$2'),
phone: (v) => v.replace(/^(.{3}).*(.{2})$/, '$1*****$2'),
ssn: (v) => `***-**-${v.slice(-4)}`,
creditCard: (v) => `****-****-****-${v.slice(-4)}`,
ip: (v) => v.replace(/\.\d+$/, '.xxx'),
name: (v) => `${v.charAt(0)}${'*'.repeat(v.length - 1)}`,
};
return Object.fromEntries(
Object.entries(data).map(([key, value]) => {
const masker = piiFields[key];
if (masker && typeof value === 'string') {
return [key, masker(value)];
}
return [key, value];
})
);
}
import { createCipheriv, createDecipheriv, randomBytes } from 'crypto';
const ALGORITHM = 'aes-256-gcm';
function encryptPII(plaintext: string, key: Buffer): EncryptedData {
const iv = randomBytes(16);
const cipher = createCipheriv(ALGORITHM, key, iv);
const encrypted = Buffer.concat([cipher.update(plaintext, 'utf8'), cipher.final()]);
const authTag = cipher.getAuthTag();
return {
ciphertext: encrypted.toString('base64'),
iv: iv.toString('base64'),
authTag: authTag.toString('base64'),
};
}
function decryptPII(data: EncryptedData, key: Buffer): string {
const decipher = createDecipheriv(
ALGORITHM,
key,
Buffer.from(data.iv, 'base64')
);
decipher.setAuthTag(Buffer.from(data.authTag, 'base64'));
return decipher.update(data.ciphertext, 'base64', 'utf8') + decipher.final('utf8');
}
0h → Breach detected
→ Incident response team activated
→ Containment measures started
24h → Impact assessment completed
→ Affected data subjects identified
→ Breach severity classified
48h → Notification to DPA prepared
→ Data subject notification prepared (if high risk)
72h → DEADLINE: DPA notification submitted (Art. 33)
→ Data subject notification sent if required (Art. 34)
interface BreachAssessment {
detectedAt: Date;
nature: string; // What happened
categoriesAffected: string[]; // Data types exposed
subjectsAffected: number; // How many people
likelyConsequences: string[]; // Potential harm
measuresTaken: string[]; // Containment actions
riskLevel: 'low' | 'medium' | 'high';
notifyDPA: boolean; // Required unless low risk
notifySubjects: boolean; // Required if high risk
dpaNotificationDeadline: Date; // detectedAt + 72h
}
function assessBreachRisk(breach: BreachAssessment): string {
const highRiskFactors = [
breach.categoriesAffected.includes('financial'),
breach.categoriesAffected.includes('health'),
breach.categoriesAffected.includes('credentials'),
breach.subjectsAffected > 1000,
];
const highRiskCount = highRiskFactors.filter(Boolean).length;
if (highRiskCount >= 2) return 'high';
if (highRiskCount >= 1) return 'medium';
return 'low';
}
interface ConsentRecord {
subjectId: string;
purpose: string; // 'marketing_email', 'analytics', 'profiling'
granted: boolean;
grantedAt: Date | null;
revokedAt: Date | null;
method: 'explicit_opt_in' | 'form' | 'api';
version: string; // Privacy policy version
ip?: string;
evidence: string; // What they agreed to (exact text)
}
async function recordConsent(consent: ConsentRecord): Promise<void> {
// Consent must be:
// - Freely given (no pre-ticked boxes)
// - Specific (per purpose)
// - Informed (clear language)
// - Unambiguous (affirmative action)
await db.consents.create({ data: consent });
await auditLog({
action: `consent.${consent.granted ? 'granted' : 'revoked'}`,
actor: consent.subjectId,
resource: consent.purpose,
details: { version: consent.version },
});
}
| Anti-Pattern | GDPR Violation | Dogru Yol |
|---|---|---|
| Pre-ticked checkboxes | Art. 7 - not freely given | Unchecked by default |
| Bundled consent | Art. 7 - not specific | Separate consent per purpose |
| Dark patterns (confusing UI) | Art. 7 - not informed | Clear, plain language |
| No withdrawal mechanism | Art. 7(3) - easy withdrawal | One-click unsubscribe |
| Consent wall (block access) | Art. 7 - not freely given | Allow access without consent |
| Implicit consent | Art. 4(11) - not unambiguous | Explicit opt-in required |
| Mechanism | When to Use | Complexity |
|---|---|---|
| Adequacy decision | EU-approved country (UK, Japan, etc.) | Low |
| Standard Contractual Clauses (SCCs) | US, India, etc. | Medium |
| Binding Corporate Rules (BCRs) | Intra-group transfers | High |
| Explicit consent | One-off transfers | Low |
| EU-US Data Privacy Framework | US companies certified | Medium |
## Transfer Impact Assessment
**From:** [EU entity]
**To:** [Non-EU entity, country]
**Mechanism:** [SCC / BCR / Adequacy / Consent]
### Data Transferred
- Categories: [personal data types]
- Volume: [approximate records/month]
- Frequency: [continuous / batch / ad-hoc]
### Recipient Country Assessment
- Privacy laws: [adequacy level]
- Government access: [surveillance risk]
- Legal remedies: [available to EU subjects?]
### Supplementary Measures
- [ ] End-to-end encryption (keys retained in EU)
- [ ] Pseudonymization before transfer
- [ ] Access controls at destination
- [ ] Contractual prohibition on government disclosure
- [ ] Regular compliance audits
| Data Category | Retention Period | Legal Basis | After Expiry |
|---|---|---|---|
| Active user data | Account lifetime | Contract | Anonymize |
| Inactive user data | 2 yil inactivity | Legitimate interest | Delete |
| Transaction records | 7 yil | Legal obligation (tax) | Archive encrypted |
| Marketing consent | Until revoked | Consent | Delete |
| Support tickets | 3 yil after resolution | Legitimate interest | Anonymize |
| Access logs | 1 yil | Legitimate interest | Delete |
| Analytics (aggregated) | Indefinite | Legitimate interest | N/A (no PII) |
async function enforceRetentionPolicies(): Promise<RetentionReport> {
const report: RetentionReport = { deletedCount: 0, anonymizedCount: 0 };
// Delete inactive accounts (2 yil)
const inactiveThreshold = subYears(new Date(), 2);
const inactiveUsers = await db.users.findMany({
where: { lastActiveAt: { lt: inactiveThreshold }, deletedAt: null },
});
for (const user of inactiveUsers) {
await executeErasure(user.id);
report.deletedCount++;
}
// Anonymize old support tickets (3 yil)
const ticketThreshold = subYears(new Date(), 3);
const oldTickets = await db.tickets.updateMany({
where: { resolvedAt: { lt: ticketThreshold }, anonymized: false },
data: { userEmail: 'anonymized', userName: 'anonymized', anonymized: true },
});
report.anonymizedCount += oldTickets.count;
// Delete old access logs (1 yil)
const logThreshold = subYears(new Date(), 1);
await db.accessLogs.deleteMany({
where: { createdAt: { lt: logThreshold } },
});
return report;
}
まだレビューはありません。使ってみた感想をお寄せください。
概要と使いどころ
WCAG 2.2 AA compliance, ARIA patterns, keyboard navigation, screen reader optimization
日本語の概要は準備中です。原文の説明を表示しています。
axe-core integration, WCAG 2.2 AA checklist, keyboard navigation testing, screen reader testing, and ARIA pattern validation.
日本語の概要は準備中です。原文の説明を表示しています。
Steam-style achievement system with XP, levels, streaks, and skill trees. Gamifies the development workflow. 25 achievements across 5 categories.
日本語の概要は準備中です。原文の説明を表示しています。
Framework for measuring and tracking agent response quality over time. Detects regressions before they reach production. Use when evaluating agent changes, auditing quality, or establishing performance baselines.
日本語の概要は準備中です。原文の説明を表示しています。
Agent Context Isolation
日本語の概要は準備中です。原文の説明を表示しています。
Agent ve skill dosyalarinin yapisal dogrulamasi. Frontmatter kontrol, naming convention, zorunlu bolum kontrolu, tutarlilik denetimi. Yeni agent/skill eklendiginde veya mevcut dosyalar duzenlediginde otomatik calistirilir.
日本語の概要は準備中です。原文の説明を表示しています。