majiayu000 Context Engineering
A practical guide for choosing, structuring, and evaluating the context supplied to LLM systems. It covers prompts, tool descriptions, retrieval, memory, compaction, autonomy, and agent architecture.
References
| # | Principle | File | Summary |
|---|---|---|---|
| 1 | Quality Over Quantity | 01-quality-over-quantity.md |
Include only relevant context. Precision beats exhaustiveness. |
| 2 | Start Minimal, Layer Progressively | 02-start-minimal.md |
Begin with essentials, expand based on performance gaps. |
| 3 | Hierarchical Context Organization | 03-hierarchical-organization.md |
Structure context from general to specific. Use XML tags for disambiguation. |
| 4 | Multi-Dimensional Context | 04-multi-dimensional-context.md |
Effective context spans task, domain, history, constraints, motivation, and format. |
| 5 | Simple Composable Patterns | 05-simple-composable-patterns.md |
Prefer prompt chaining, routing, parallelization over complex frameworks. |
| 6 | Invest in Tool Design (ACI) | 06-invest-in-tool-design.md |
Tool descriptions deserve as much effort as UIs. Calibrate action language for Claude 4.6. |
| 7 | Match Architecture to Complexity | 07-match-architecture-to-complexity.md |
Choose the right pattern: chaining, routing, parallelization, orchestrator, or agent. |
| 8 | Transparency in Agent Planning | 08-transparency-in-planning.md |
Show planning steps. Use structured state tracking for long-horizon tasks. |
| 9 | Contextualize Retrieved Chunks | 09-contextualize-retrieved-chunks.md |
Prepend explanatory context before embedding. Reduces retrieval failures by 35-67%. |
| 10 | Dynamic Context Adaptation | 10-dynamic-context-adaptation.md |
Adjust context based on task complexity and model confidence. |
| 11 | Meta-Context | 11-meta-context.md |
Explain why information is provided. Use grounding directives to reduce hallucinations. |
| 12 | Ground Agents in Environment Feedback | 12-ground-in-environment.md |
Use tool results and human checkpoints, not just model reasoning. |
| 13 | Test and Measure Systematically | 13-test-and-measure.md |
A/B test context variations. Measure accuracy, latency, reasoning quality. |
| 14 | Retrieve Generously, Filter Late | 14-retrieve-generously.md |
Top-20 chunks outperform top-5. Don't over-filter prematurely. |
| 15 | Avoid Common Anti-Patterns | 15-anti-patterns.md |
Don't dump everything, over-abstract, overengineer, or inherit stale aggressive prompts. |
| 16 | Guide Model Thinking | 16-guide-model-thinking.md |
Control reasoning depth with adaptive thinking. Prevent overthinking on simple tasks. |
| 17 | Multi-Window State Management | 17-multi-window-state.md |
Externalize state for tasks spanning multiple context windows. |
| 18 | Calibrate Autonomy and Safety | 18-calibrate-autonomy.md |
Tune action thresholds. Claude 4.6 defaults to high autonomy; set guardrails. |