Skip to content
DKKB
Search
Ctrl
K
Cancel
GitHub
Select theme
Dark
Light
Auto
Foundations
Principles
Overview
Prefer composition over inheritance
Invert dependencies around policy
Do not repeat knowledge
Keep monetization out of relevance ranking
Separate concerns
You aren't gonna need it
Architecture
Overview
Architecture boundaries and dependency direction
Event-driven architecture
Layered architecture
Modular monolith
Ports and Adapters
Decisions
Overview
Anonymous-first versus account-required product design
Separate public and operator surfaces
Patterns
Overview
Adapter pattern
Separate canonical domain entities from provider representations
Observer pattern
Design provider-neutral integration boundaries
Separate current state from observation history
Strategy
Anti-patterns
Overview
God object
Shotgun surgery
Problems
Overview
N+1 query problem
Race condition
Retry storm
Thundering herd
Practices
Overview
Conventional Commits
Gate data publication by explicit rights state
Preserve provenance during canonicalization
Safe online data migrations
Engineering
Coding
Overview
Control hidden side effects and temporal coupling
Keep functions and modules cohesive
Make invariants explicit at boundaries
Name for intent and domain meaning
Refactor in behavior-preserving steps
Testing
Overview
Contract testing
Deterministic tests
Test doubles
Testing strategy
Unit, integration, and end-to-end tests
Reliability
Overview
Circuit breakers
Idempotency
Load shedding and backpressure
Retries and exponential backoff
Timeouts
Observability
Overview
Design actionable alerts
Logs, metrics, and traces
Debug production behavior with evidence
Service-level indicators and objectives
Structured logging and correlation identifiers
Performance
Overview
Bound candidate retrieval before expensive ranking
Bounded work
Caching
Latency vs throughput
Measure before optimizing
Performance budgets and profiling
Security
Overview
Authentication vs authorization
Defense in depth
Least privilege
Secure defaults and fail-closed behavior
Threat modeling
Databases
Overview
Indexes and query planning
Isolation levels and concurrency anomalies
Normalization versus denormalization
Optimistic versus pessimistic concurrency control
Transactions and consistency boundaries
API design
Overview
API contracts and compatibility
API versioning and evolution
Error modeling and error normalization
Pagination and bounded collection reads
Resource and operation design
AI
AI engineering
Overview
Bound AI agents by authority and verification
Bound AI implementation work to one verifiable unit
LLM engineering
Overview
Context engineering for LLM systems
Evaluate LLM behavior and hallucination risk
Retrieval-augmented generation
Structured LLM output
Personal
Playbook
Overview
Literature ingestion workflow
Personal experience capture workflow
Reference
Glossary
Overview
References
Overview
GitHub
Select theme
Dark
Light
Auto
Glossary
Section titled “Glossary”
Definitions for terms that need a stable meaning across DKKB.