Engineering Prompts

---

LEVEL 1: QUICKSTART ⚡

Foundation First (Always Apply):

1. Clarity: Explicit instructions + success criteria
2. Context & Motivation: Explain WHY, not just WHAT
3. Positive Framing: Say what TO do, not what NOT to do

Then Assess:

4. Simple task? → Stop here. Foundation is enough.
5. Complex task? → Add advanced techniques sparingly
6. Test & Iterate: Measure effectiveness, refine based on results

Key insight: The best prompt achieves goals with minimum necessary structure.

---

LEVEL 2: DIAGNOSTICS 🔧

Output Problems → Solutions:

Problem	Solution
Too generic	Add specificity + examples
Off-topic	Add context explaining goals
Wrong format	Examples or prefilling
Unreliable on complex tasks	Prompt chaining
Unnecessary preambles	Prefill or "start directly with..."
Hallucinations	"If unsure, acknowledge uncertainty"
Shallow reasoning	Chain of thought

Process Problems → Fixes:

Mistake	Fix
Over-engineering	Start minimal, add only what helps
Negative framing	Say what TO do instead
No motivation	Explain the WHY
No iteration	Test, measure, refine

---

LEVEL 3: THE 12 TECHNIQUES 🛠️

Foundation (Always Apply)

Technique	What It Does	Cost
1. Clarity	Explicit instructions, success criteria	Minimal
2. Context & Motivation	Explain WHY you want something	Minimal
3. Positive Framing	Say what TO do, not what NOT to do	Minimal
4. XML Structure	Separate sections in complex prompts	~50-100 tokens

Note: XML is less essential with Claude 4.x - use when genuinely needed.

Advanced (Apply When Needed)

Technique	When to Use	Cost
5. Chain of Thought	Reasoning, analysis, math	2-3x output
6. Prompt Chaining	Multi-step, complex tasks	Multiple calls
7. Multishot Examples	Pattern learning, format	200-1K each
8. System Role	Domain expertise needed	Minimal
9. Prefilling	Strict format requirements	Minimal
10. Long Context	20K+ token inputs	Better accuracy
11. Context Budget	Repeated use, conversations	90% savings
12. Tool Docs	Function calling, agents	100-500/tool

---

LEVEL 4: DESIGN CHECKLIST 📋

Before Writing:

• Core task clear?
• Output format defined?
• Constraints known? (latency/cost/accuracy)
• One-off or repeated use?

Complexity Assessment:

• Simple (extraction, Q&A) → Foundation only
• Medium (analysis, code gen) → + CoT, examples
• Complex (research, novel problems) → + Chaining, role

Cost Optimization:

• Cache system prompts + reference docs (90% savings)
• Batch non-urgent work (50% savings)
• Skip CoT for simple tasks (saves 2-3x)

Deliverable:

• Prompt + techniques used + rationale + token estimate

---

LEVEL 5: ADVANCED TOPICS 🚀

Tool Integration

When to use MCP tools during prompt engineering:

Need latest practices?
└─ mcp__plugin_essentials_perplexity

Complex analysis needed?
└─ mcp__plugin_essentials_sequential-thinking

Need library docs?
└─ mcp__plugin_essentials_context7

Context Management

Prompt Caching:

• Cache: System prompts, reference docs, examples
• Savings: 90% on cached content
• Write: 25% of standard cost
• Read: 10% of standard cost

Long Context Tips:

• Place documents BEFORE queries
• Use XML tags: <document>, <source>
• Ground responses in quotes
• 30% better performance with proper structure

Token Optimization

Reducing Token Usage:

• Concise, clear instructions (no fluff)
• Reuse examples across calls (cache them)
• Structured output reduces back-and-forth
• Tool use instead of long context when possible

Cost-Specific Anti-Patterns

❌ Ignoring caching - Not leveraging repeated content (90% savings lost) ❌ Over-requesting CoT - Chain of thought for simple tasks (2-3x wasted) ❌ Redundant examples - 5 examples when 2 suffice ❌ No batching - Real-time calls for non-urgent work (50% savings lost)

See LEVEL 2: Diagnostics for general fixes.

---

LEVEL 6: REFERENCES 📚

Deep Dive Documentation

Detailed Technique Catalog:

• reference/technique-catalog.md - Each technique explained with examples, token costs, combination strategies

Real-World Examples:

• reference/examples.md - Before/after pairs for coding, analysis, extraction, agent tasks

Research Papers:

• reference/research.md - Latest Anthropic research, benchmarks, best practices evolution