context-engineering - Context Engineering for Token Optimization
Maximizes agent output quality while minimizing token expenditure through relevance scoring, context compression, budget allocation, and context reclamation.
Tags
Updated: 2026-09-20Capabilities
Typical Inputs
Typical Outputs
What this skill does
- Score relevance of context sources
- Compress conversation history via summarization
- Enforce context token budget limits
- Apply progressive context disclosure layers
- Reclaim unused tool output context
Inputs
- Raw context sources
- Token budget parameters
- Task specifications
Outputs
- Assembled context window
- Compressed context summary
- Session memory updates
Requirements
- Compatible agent harness
- Token counting capability