Context Optimization

Context Optimization Techniques

Context optimization extends the effective capacity of limited context windows through strategic compression, masking, caching, and partitioning. Effective optimization can double or triple effective context capacity without requiring larger models or longer windows — but only when applied with discipline. The techniques below are ordered by impact and risk.

When to Activate

Activate this skill when:

• Context limits constrain task complexity
• Optimizing for cost reduction (fewer tokens = lower costs)
• Reducing latency for long conversations
• Implementing long-running agent systems
• Needing to handle larger documents or conversations
• Building production systems at scale

Core Concepts

Apply four primary strategies in this priority order:

1. KV-cache optimization — Reorder and stabilize prompt structure so the inference engine reuses cached Key/Value tensors. This is the cheapest optimization: zero quality risk, immediate cost and latency savings. Apply it first and unconditionally.

2. Observation masking — Replace verbose tool outputs with compact references once their purpose has been served. Tool outputs consume 80%+ of tokens in typical agent trajectories, so masking them yields the largest capacity gains. The original content remains retrievable if needed downstream.

3. Compaction — Summarize accumulated context when utilization exceeds 70%, then reinitialize with the summary. This distills the window's contents while preserving task-critical state. Compaction is lossy — apply it after masking has already removed the low-value bulk.

4. Context partitioning — Split work across sub-agents with isolated contexts when a single window cannot hold the full problem. Each sub-agent operates in a clean context focused on its subtask. Reserve this for tasks where estimated context exceeds 60% of the window limit, because coordination overhead is real.

The governing principle: context quality matters more than quantity. Every optimization preserves signal while reducing noise. Measure before optimizing, then measure the optimization's effect.

Detailed Topics

Compaction Strategies

Trigger compaction when context utilization exceeds 70%: summarize the current context, then reinitialize with the summary. This distills the window's contents in a high-fidelity manner, enabling continuation with minimal performance degradation. Prioritize compressing tool outputs first (they consume 80%+ of tokens), then old conversation turns, then retrieved documents. Never compress the system prompt — it anchors model behavior and its removal causes unpredictable degradation.

Preserve different elements by message type:

• Tool outputs: Extract key findings, metrics, error codes, and conclusions. Strip verbose raw output, stack traces (unless debugging is ongoing), and boilerplate headers.
• Conversational turns: Retain decisions, commitments, user preferences, and context shifts. Remove filler, pleasantries, and exploratory back-and-forth that led to a conclusion already captured.
• Retrieved documents: Keep claims, facts, and data points relevant to the active task. Remove supporting evidence and elaboration that served a one-time reasoning purpose.

Target 50-70% token reduction with less than 5% quality degradation. If compaction exceeds 70% reduction, audit the summary for critical information loss — over-aggressive compaction is the most common failure mode.

Observation Masking

Mask observations selectively based on recency and ongoing relevance — not uniformly. Apply these rules:

• Never mask: Observations critical to the current task, observations from the most recent turn, observations used in active reasoning chains, and error outputs when debugging is in progress.
• Mask after 3+ turns: Verbose outputs whose key points have already been extracted into the conversation flow. Replace with a compact reference: [Obs:{ref_id} elided. Key: {summary}. Full content retrievable.]
• Always mask immediately: Repeated/duplicate outputs, boilerplate headers and footers, outputs already summarized earlier in the conversation.

Masking should achieve 60-80% reduction in masked observations with less than 2% quality impact. The key is maintaining retrievability — store the full content externally and keep the reference ID in context so the agent can request the original if needed.

KV-Cache Optimization

Maximize prefix cache hits by structuring prompts so that stable content occupies the prefix and dynamic content appears at the end. KV-cache stores Key and Value tensors computed during inference; when consecutive requests share an identical prefix, the cached tensors are reused, saving both cost and latency.

Apply this ordering in every prompt: 1. System prompt (most stable — never changes within a session) 2. Tool definitions (stable across requests) 3. Frequently reused templates and few-shot examples 4. Conversation history (grows but shares prefix with prior turns) 5. Current query and dynamic content (least stable — always last)

Design prompts for cache stability: remove timestamps, session counters, and request IDs from the system prompt. Move dynamic metadata into a separate user message or tool result where it does not break the prefix. Even a single whitespace change in the prefix invalidates the entire cached block downstream of that change.

Target 70%+ cache hit rate for stable workloads. At scale, this translates to 50%+ cost reduction and 40%+ latency reduction on cached tokens.

Context Partitioning

Partition work across sub-agents when a single context cannot hold the full problem without triggering aggressive compaction. Each sub-agent operates in a clean, focused context for its subtask, then returns a structured result to a coordinator agent.

Plan partitioning when estimated task context exceeds 60% of the window limit. Decompose the task into independent subtasks, assign each to a sub-agent, and aggregate results. Validate that all partitions completed before merging, merge compatible results, and apply summarization if the aggregated output still exceeds budget.

This approach achieves separation of concerns — detailed search context stays isolated within sub-agents while the coordinator focuses on synthesis. However, coordination has real token cost: the coordinator prompt, result aggregation, and error handling all consume tokens. Only partition when the savings exceed this overhead.

Budget Management

Allocate explicit token budgets across context categories before the session begins: system prompt, tool definitions, retrieved documents, message history, tool outputs, and a reserved buffer (5-10% of total). Monitor usage against budget continuously and trigger optimization when any category exceeds its allocation or total utilization crosses 70%.

Use trigger-based optimization rather than periodic optimization. Monitor these signals:

• Token utilization above 80% — trigger compaction
• Attention degradation indicators (repetition, missed instructions) — trigger masking + compaction
• Quality score drops below baseline — audit context composition before optimizing

Practical Guidance

Optimization Decision Framework

Select the optimization technique based on what dominates the context:

Performance Targets

Track these metrics to validate optimization effectiveness:

• Compaction: 50-70% token reduction, <5% quality degradation, <10% latency overhead from the compaction step itself
• Masking: 60-80% reduction in masked observations, <2% quality impact, near-zero latency overhead
• Cache optimization: 70%+ hit rate for stable workloads, 50%+ cost reduction, 40%+ latency reduction
• Partitioning: Net token savings after accounting for coordinator overhead; break-even typically requires 3+ subtasks

Iterate on strategies based on measured results. If an optimization technique does not measurably improve the target metric, remove it — optimization machinery itself consumes tokens and adds latency.

Examples

Example 1: Compaction Trigger

if context_tokens / context_limit > 0.8:
    context = compact_context(context)

Example 2: Observation Masking

if len(observation) > max_length:
    ref_id = store_observation(observation)
    return f"[Obs:{ref_id} elided. Key: {extract_key(observation)}]"

Example 3: Cache-Friendly Ordering

# Stable content first
context = [system_prompt, tool_definitions]  # Cacheable
context += [reused_templates]  # Reusable
context += [unique_content]  # Unique

Guidelines

1. Measure before optimizing—know your current state 2. Apply masking before compaction — remove low-value bulk first, then summarize what remains 3. Design for cache stability with consistent prompts 4. Partition before context becomes problematic 5. Monitor optimization effectiveness over time 6. Balance token savings against quality preservation 7. Test optimization at production scale 8. Implement graceful degradation for edge cases

Gotchas

1. Whitespace breaks KV-cache: Even a single whitespace or newline change in the prompt prefix invalidates the entire KV-cache block downstream of that point. Pin system prompts as immutable strings — do not interpolate timestamps, version numbers, or session IDs into them. Diff prompt templates byte-for-byte between deployments.

2. Timestamps in system prompts destroy cache hit rates: Including Current date: {today} or similar dynamic content in the system prompt forces a full cache miss on every new day (or every request, if using time-of-day). Move dynamic metadata into a user message or a separate tool result appended after the stable prefix.

3. Compaction under pressure loses critical state: When the model performing compaction is itself under context pressure (>85% utilization), its summarization quality degrades — it omits task goals, drops user constraints, and flattens nuanced state. Trigger compaction at 70-80%, not 90%+. If compaction must happen late, use a separate model call with a clean context containing only the material to summarize.

4. Masking error outputs breaks debugging loops: Over-aggressive masking hides error messages, stack traces, and failure details that the agent needs in subsequent turns to diagnose and fix issues. During active debugging (error in the last 3 turns), suspend masking for all error-related observations until the issue is resolved.

5. Partitioning overhead can exceed savings: Each sub-agent requires its own system prompt, tool definitions, and coordination messages. For tasks with fewer than 3 independent subtasks, the coordination overhead often exceeds the context savings. Estimate total tokens (coordinator + all sub-agents) before committing to partitioning.

6. Cache miss cost spikes after deployment changes: Reordering tools, rewording the system prompt, or changing few-shot examples between deployments invalidates the entire prefix cache, causing a temporary cost spike of 2-5x until the new cache warms up. Roll out prompt changes gradually and monitor cache hit rate during deployment windows.

7. Compaction creates false confidence in stale summaries: Once context is compacted, the summary looks authoritative but may reflect outdated state. If the task has evolved since compaction (new user requirements, corrected assumptions), the summary silently carries forward stale information. After compaction, re-validate the summary against the current task goal before proceeding.

Integration

This skill builds on context-fundamentals and context-degradation. It connects to:

• multi-agent-patterns - Partitioning as isolation
• latent-briefing - Selective KV retention across orchestrator–worker boundaries (compatible models)
• evaluation - Measuring optimization effectiveness
• memory-systems - Offloading context to memory

References

Internal reference:

• Optimization Techniques Reference - Read when: implementing a specific optimization technique and needing detailed code patterns, threshold tables, or integration examples beyond what the skill body provides

Related skills in this collection:

• context-fundamentals - Read when: unfamiliar with context window mechanics, token counting, or attention distribution basics
• context-degradation - Read when: diagnosing why agent performance has dropped and needing to identify which degradation pattern is occurring before selecting an optimization
• evaluation - Read when: setting up metrics and benchmarks to measure whether an optimization technique actually improved outcomes

External resources:

• Research on context window limitations - Read when: evaluating model-specific context behavior (e.g., lost-in-the-middle effects, attention decay curves)
• KV-cache optimization techniques - Read when: implementing prefix caching at the inference infrastructure level (vLLM, TGI, or cloud provider APIs)
• Production engineering guides - Read when: deploying context optimization in a production pipeline and needing operability patterns (monitoring, alerting, rollback)

---

Skill Metadata

Created: 2025-12-20 Last Updated: 2026-03-17 Author: Agent Skills for Context Engineering Contributors Version: 2.0.0

Context Optimization Reference

This document provides detailed technical reference for context optimization techniques and strategies.

Compaction Strategies

Summary-Based Compaction

Summary-based compaction replaces verbose content with concise summaries while preserving key information. The approach works by identifying sections that can be compressed, generating summaries that capture essential points, and replacing full content with summaries.

The effectiveness of compaction depends on what information is preserved. Critical decisions, user preferences, and current task state should never be compacted. Intermediate results and supporting evidence can be summarized more aggressively. Boilerplate, repeated information, and exploratory reasoning can often be removed entirely.

Token Budget Allocation

Effective context budgeting requires understanding how different context components consume tokens and allocating budget strategically:

Compaction Thresholds

Trigger compaction at appropriate thresholds to maintain performance:

• Warning threshold at 70% of effective context limit
• Compaction trigger at 80% of effective context limit
• Aggressive compaction at 90% of effective context limit

The exact thresholds depend on model behavior and task characteristics. Some models show graceful degradation while others exhibit sharp performance cliffs.

Observation Masking Patterns

Selective Masking

Not all observations should be masked equally. Consider masking observations that have served their purpose and are no longer needed for active reasoning. Keep observations that are central to the current task. Keep observations from the most recent turn. Keep observations that may be referenced again.

Masking Implementation

def selective_mask(observations: List[Dict], current_task: Dict) -> List[Dict]:
    """
    Selectively mask observations based on relevance.
    
    Returns observations with mask field indicating masked content.
    """
    masked = []
    
    for obs in observations:
        relevance = calculate_relevance(obs, current_task)
        
        if relevance < 0.3 and obs["age"] > 3:
            # Low relevance and old - mask
            masked.append({
                **obs,
                "masked": True,
                "reference": store_for_reference(obs["content"]),
                "summary": summarize_content(obs["content"])
            })
        else:
            masked.append({
                **obs,
                "masked": False
            })
    
    return masked

KV-Cache Optimization

Prefix Stability

KV-cache hit rates depend on prefix stability. Stable prefixes enable cache reuse across requests. Dynamic prefixes invalidate cache and force recomputation.

Elements that should remain stable include system prompts, tool definitions, and frequently used templates. Elements that may vary include timestamps, session identifiers, and query-specific content.

Cache-Friendly Design

Design prompts to maximize cache hit rates:

1. Place stable content at the beginning 2. Use consistent formatting across requests 3. Avoid dynamic content in prompts when possible 4. Use placeholders for dynamic content

# Cache-unfriendly: Dynamic timestamp in prompt
system_prompt = f"""
Current time: {datetime.now().isoformat()}
You are a helpful assistant.
"""

# Cache-friendly: Stable prompt with dynamic time as variable
system_prompt = """
You are a helpful assistant.
Current time is provided separately when relevant.
"""

Context Partitioning Strategies

Sub-Agent Isolation

Partition work across sub-agents to prevent any single context from growing too large. Each sub-agent operates with a clean context focused on its subtask.

Partition Planning

def plan_partitioning(task: Dict, context_limit: int) -> Dict:
    """
    Plan how to partition a task based on context limits.
    
    Returns partitioning strategy and subtask definitions.
    """
    estimated_context = estimate_task_context(task)
    
    if estimated_context <= context_limit:
        return {
            "strategy": "single_agent",
            "subtasks": [task]
        }
    
    # Plan multi-agent approach
    subtasks = decompose_task(task)
    
    return {
        "strategy": "multi_agent",
        "subtasks": subtasks,
        "coordination": "hierarchical"
    }

Optimization Decision Framework

When to Optimize

Consider context optimization when context utilization exceeds 70%, when response quality degrades as conversations extend, when costs increase due to long contexts, or when latency increases with conversation length.

What Optimization to Apply

Choose optimization strategies based on context composition:

If tool outputs dominate context, apply observation masking. If retrieved documents dominate context, apply summarization or partitioning. If message history dominates context, apply compaction with summarization. If multiple components contribute, combine strategies.

Evaluation of Optimization

After applying optimization, evaluate effectiveness:

• Measure token reduction achieved
• Measure quality preservation (output quality should not degrade)
• Measure latency improvement
• Measure cost reduction

Iterate on optimization strategies based on evaluation results.

Common Pitfalls

Over-Aggressive Compaction

Compacting too aggressively can remove critical information. Always preserve task goals, user preferences, and recent conversation context. Test compaction at increasing aggressiveness levels to find the optimal balance.

Masking Critical Observations

Masking observations that are still needed can cause errors. Track observation usage and only mask content that is no longer referenced. Consider keeping references to masked content that could be retrieved if needed.

Ignoring Attention Distribution

The lost-in-middle phenomenon means that information placement matters. Place critical information at attention-favored positions (beginning and end of context). Use explicit markers to highlight important content.

Premature Optimization

Not all contexts require optimization. Adding optimization machinery has overhead. Optimize only when context limits actually constrain agent performance.

Monitoring and Alerting

Key Metrics

Track these metrics to understand optimization needs:

• Context token count over time
• Cache hit rates for repeated patterns
• Response quality metrics by context size
• Cost per conversation by context length
• Latency by context size

Alert Thresholds

Set alerts for:

• Context utilization above 80%
• Cache hit rate below 50%
• Quality score drop of more than 10%
• Cost increase above baseline

Integration Patterns

Integration with Agent Framework

Integrate optimization into agent workflow:

class OptimizingAgent:
    def __init__(self, context_limit: int = 80000):
        self.context_limit = context_limit
        self.optimizer = ContextOptimizer()
    
    def process(self, user_input: str, context: Dict) -> Dict:
        # Check if optimization needed
        if self.optimizer.should_compact(context):
            context = self.optimizer.compact(context)
        
        # Process with optimized context
        response = self._call_model(user_input, context)
        
        # Track metrics
        self.optimizer.record_metrics(context, response)
        
        return response

Integration with Memory Systems

Connect optimization with memory systems:

class MemoryAwareOptimizer:
    def __init__(self, memory_system, context_limit: int):
        self.memory = memory_system
        self.limit = context_limit
    
    def optimize_context(self, current_context: Dict, task: str) -> Dict:
        # Check if information is in memory
        relevant_memories = self.memory.retrieve(task)
        
        # Move information to memory if not needed in context
        for mem in relevant_memories:
            if mem["importance"] < threshold:
                current_context = remove_from_context(current_context, mem)
                # Keep reference that memory can be retrieved
        
        return current_context

Performance Benchmarks

Compaction Performance

Compaction should reduce token count while preserving quality. Target:

• 50-70% token reduction for aggressive compaction
• Less than 5% quality degradation from compaction
• Less than 10% latency increase from compaction overhead

Masking Performance

Observation masking should reduce token count significantly:

• 60-80% reduction in masked observations
• Less than 2% quality impact from masking
• Near-zero latency overhead

Cache Performance

KV-cache optimization should improve cost and latency:

• 70%+ cache hit rate for stable workloads
• 50%+ cost reduction from cache hits
• 40%+ latency reduction from cache hits