claudekit/skillmark/ck-context-engineering

Context Engineering

Context engineering curates the smallest high-signal token set for LLM tasks. The goal: maximize reasoning quality while minimizing token usage.

When to Activate

• Designing/debugging agent systems
• Context limits constrain performance
• Optimizing cost/latency
• Building multi-agent coordination
• Implementing memory systems
• Evaluating agent performance
• Developing LLM-powered pipelines

Core Principles

1. Context quality > quantity - High-signal tokens beat exhaustive content
2. Attention is finite - U-shaped curve favors beginning/end positions
3. Progressive disclosure - Load information just-in-time
4. Isolation prevents degradation - Partition work across sub-agents
5. Measure before optimizing - Know your baseline

IMPORTANT:

• Sacrifice grammar for the sake of concision.
• Ensure token efficiency while maintaining high quality.
• Pass these rules to subagents.

Quick Reference

Key Concepts Index

• Multi-agent: orchestrator/supervisor patterns, consensus mechanisms, voting, 15x token economics, context isolation, handoff protocols, specialization
• Optimization: compaction triggers (70-80%), observation masking (tool outputs), KV-cache/prefix caching, progressive disclosure, priority preservation
• Memory: short-term (session), long-term (persistent), RAG retrieval, vector embeddings, knowledge graphs, cross-session persistence
• Evaluation: token metrics, quality metrics, LLM-as-Judge, probe-based testing, baseline comparison, degradation detection

Key Metrics

• Token utilization: Warning at 70%, trigger optimization at 80%
• Token variance: Explains 80% of agent performance variance
• Multi-agent cost: ~15x single agent baseline
• Compaction target: 50-70% reduction, <5% quality loss
• Cache hit target: 70%+ for stable workloads

Four-Bucket Strategy

1. Write: Save context externally (scratchpads, files)
2. Select: Pull only relevant context (retrieval, filtering)
3. Compress: Reduce tokens while preserving info (summarization)
4. Isolate: Split across sub-agents (partitioning)

Anti-Patterns

• Exhaustive context over curated context
• Critical info in middle positions
• No compaction triggers before limits
• Single agent for parallelizable tasks
• Tools without clear descriptions

Guidelines

1. Place critical info at beginning/end of context
2. Implement compaction at 70-80% utilization
3. Use sub-agents for context isolation, not role-play
4. Design tools with 4-question framework (what, when, inputs, returns)
5. Optimize for tokens-per-task, not tokens-per-request
6. Validate with probe-based evaluation
7. Monitor KV-cache hit rates in production
8. Start minimal, add complexity only when proven necessary

Runtime Awareness

The system automatically injects usage awareness via PostToolUse hook:

<usage-awareness>
Claude Usage Limits: 5h=45%, 7d=32%
Context Window Usage: 67%
</usage-awareness>

Thresholds:

• 70%: WARNING - consider optimization/compaction
• 90%: CRITICAL - immediate action needed

Data Sources:

• Usage limits: Anthropic OAuth API (https://api.anthropic.com/api/oauth/usage)
• Context window: Statusline temp file (/tmp/ck-context-{session_id}.json)

Scripts

• context_analyzer.py - Context health analysis, degradation detection
• compression_evaluator.py - Compression quality evaluation