Medium LLM
30 Prompt Techniques I Actually Use With Claude (Not a Copy-Paste Listicle)
context
What happened
The author shares practical prompting techniques for Claude based on real-world usage. Key highlights include using explicit constraints (e.g., defining target audience and length), utilizing XML-style tags to structure prompts (separating instructions, context, and examples), and building systemic workflows rather than single-shot prompts.
Why it matters
Well-structured prompts with clear boundaries (like XML tags) remain the baseline for reliable LLM tool use and agent instructions.
The take
While prompt engineering lists are common, emphasizing XML tags and structural separation is indeed best practice for Anthropic's models. It's a solid refresher on prompt hygiene, though it lacks deep architectural novelty.
Do this
Adopt XML tags systematically in your Claude prompts to cleanly separate system instructions, user inputs, and few-shot examples.
Don't read this site daily. Get it in your inbox.
The daily brief and Sunday deep dive — distilled, scored, and opinionated. For builders only.