SkillHack › Career guides › Prompt Engineer
How to become an Prompt Engineer
Prompt design, context engineering, and eval-driven iteration. Get reliable, repeatable behavior out of frontier models.
- 5phases in the roadmap
- 14topics to work through
- 15graded practice questions
- Freeno payment, ever
The Prompt Engineer roadmap
Each topic below carries a study note, real reading, and practice questions in the app. A topic unlocks when you pass the one before it, so the order is the path, so you are never guessing what to learn next.
Phase 1Prompt engineering prompting foundations
Anatomy of a prompt
A prompt is the model's source code, and structure is what makes it followable.
Few-shot examples
Examples teach a model the pattern you want faster than description alone.
Knowing what not to ask the model
Half of prompt engineering is knowing the model's limits: it is the wrong tool for exact arithmetic, fresh facts, and deterministic lookups.
Phase 2Reasoning and output control
Chain-of-thought and when it hurts
Asking a model to reason step by step helps on genuinely multi-step problems - and hurts on simple ones, where it adds latency, cost, and chances to talk itself into a wrong answer.
Structured and constrained output
Agents and pipelines consume model output programmatically, so free-form prose is a bug surface.
Sampling controls
Temperature, top-p, and stop sequences decide how deterministic or creative the output is.
Phase 3Prompt engineering grounding and citations
Grounding and citation requirements
Retrieval-augmented prompts constrain the model to answer from supplied sources and cite them.
Context budgeting
Context is a finite budget split between instructions, retrieved material, history, and the answer.
Phase 4Testing and reliability
Evaluating and iterating on prompts
A prompt that looks better on two playground examples may have regressed ten you did not check.
Diagnosing malformed output
Truncated JSON, unescaped characters, and format drift are everyday failures once output is consumed by code.
Prompt versioning and management
Prompts scattered as string literals become un-diffable and un-rollback-able the moment one causes a regression.
Phase 5Prompt engineering production prompting
Prompt injection defenses
Any text a prompt incorporates - a web page, an email, a document - can carry instructions that hijack the model.
Cost and latency of prompts
Every token in a prompt is paid for on every call, so a bloated prompt is a recurring tax on latency and cost.
System prompts and role design
Role, tone, and refusal policy belong in a stable system prompt - persistent instructions applied every turn and kept separate from user input.
Reading for this path
The primary sources behind the topics above, all free to read.
- Anthropic: prompt engineering overview
- Anthropic: use XML tags
- Anthropic: multishot prompting
- Anthropic: reduce hallucinations
- Anthropic: chain-of-thought
- Chain-of-Thought prompting (paper)
- Anthropic: increase output consistency
- Anthropic: Messages API (sampling parameters)
- Anthropic: long context tips
- Anthropic: create strong empirical evaluations
- OWASP Top 10 for LLM Applications
- Anthropic: reduce prompt injection
- Anthropic: system prompts
Coming from another job?
Most people on this path arrived from somewhere else: engineering, analysis, testing, product, support, compliance, design. Onboarding asks what you do today and builds a short starter run out of the gaps, so you begin from what you already know rather than from chapter one. See how it works.
Other AI career paths
Recommender Systems Engineer
Candidate generation, ranking models, embeddings, and feedback loops. Personalize what every user sees at scale.
Robotics / Embodied AI Engineer
Perception, control, simulation, and vision-language-action models. Put AI to work in the physical world.
Search & Ranking Engineer
Classic information retrieval, neural retrieval, and LLM re-ranking. Return the right result first, at scale.
Synthetic Data Engineer
Data generation, augmentation, and privacy-preserving datasets. Manufacture the data real-world collection can't provide.
Voice AI Engineer
Speech recognition, text-to-speech, voice agents, and real-time audio. Build AI you can talk to.
Agent Engineer
Multi-agent architectures, orchestration, tool design, and agent reliability. Build agent systems that plan, act, and recover on their own.
Start the Prompt Engineer path for free
The full study notes, the reading, and the practice questions behind every topic above are in the app. Answer one tonight and you have started.
No payment, no credit card, no CV. Sign in with Google.