Atlas · GenAI 2026
Prompt Engineering & Model Interaction
21 skills · ontology graph below shows relations within this section.
What this domain covers
This edition groups 21 capabilities in Prompt Engineering & Model Interaction across 13 named categories. The inventory contains 17 concepts and 4 tools; 8 skills appeared in at least two of the three original research runs. The remaining entries stay visible with their lower coverage so a reader can distinguish taxonomy scope from research-system agreement.
Current category labels: Context Engineering · Data Generation · Decoding · Inference Efficiency · LLM APIs & SDKs · Model Access · Model Routing · Prompt Design · and 5 more
Frequent learning foundations
- Prompt Engineering supports 8 mapped skills
- Context Engineering supports 2 mapped skills
- LLM Decoding Strategies supports 2 mapped skills
- Transformer Architecture supports 2 mapped skills
- A/B Testing supports 1 mapped skill
Skills in this section
Context Engineering
Context Engineering
Synthetic Data Generation
Data Generation
LLM Decoding Strategies
Decoding
Prompt Caching
Inference Efficiency
Token Optimization
Inference Efficiency
Anthropic API
LLM APIs & SDKs
OpenAI API
LLM APIs & SDKs
LLM API Integration
Model Access
Semantic Routing
Model Routing
In-Context Learning
Prompt Design
Prompt Engineering
Prompt Design
System Prompt Design
Prompt Design
Prompt Management
Prompt Ops
Automated Prompt Optimization
Prompt Optimization Frameworks
DSPy
Prompt Optimization Frameworks
Program-Aided LMs (PAL)
Reasoning Techniques
Self-Consistency
Reasoning Techniques
Structured LLM Outputs
Structured Outputs
Test-Time Compute Scaling
Test-Time Compute
Google Gemini API
LLM APIs & SDKs
Chain-of-Thought Prompting
Reasoning Techniques