13 KiB
13 KiB
Memory-Enhanced Context Restoration Task
Purpose
Intelligently restore context using both session state and accumulated memory insights to provide comprehensive, actionable context for persona activation and task execution.
Multi-Layer Context Restoration Process
1. Session State Analysis
Immediate Context Loading:
def load_session_context():
session_state = load_file('.ai/orchestrator-state.md')
return {
"project_name": extract_project_name(session_state),
"current_phase": extract_current_phase(session_state),
"active_personas": extract_persona_history(session_state),
"recent_decisions": extract_decision_log(session_state),
"pending_items": extract_blockers_and_concerns(session_state),
"last_activity": extract_last_activity(session_state),
"session_duration": calculate_session_duration(session_state)
}
2. Memory Intelligence Integration
Historical Context Queries:
def gather_memory_intelligence(session_context, target_persona):
memory_queries = []
# Direct persona relevance
memory_queries.append(f"{target_persona} successful patterns {session_context.project_type}")
# Current phase insights
memory_queries.append(f"{session_context.current_phase} challenges solutions {target_persona}")
# Pending items resolution
if session_context.pending_items:
memory_queries.append(f"solutions for {session_context.pending_items}")
# Cross-project learning
memory_queries.append(f"successful {target_persona} approaches {session_context.tech_context}")
# Anti-pattern prevention
memory_queries.append(f"common mistakes {target_persona} {session_context.current_phase}")
return execute_memory_queries(memory_queries)
def execute_memory_queries(queries):
memory_insights = {
"relevant_patterns": [],
"success_strategies": [],
"anti_patterns": [],
"optimization_opportunities": [],
"personalization_insights": []
}
for query in queries:
memories = search_memory(query, limit=3, threshold=0.7)
categorize_memories(memories, memory_insights)
return memory_insights
3. Proactive Intelligence Generation
Intelligent Anticipation:
def generate_proactive_insights(session_context, memory_insights, target_persona):
proactive_intelligence = {}
# Predict likely next actions
proactive_intelligence["likely_next_actions"] = predict_next_actions(
session_context, memory_insights, target_persona
)
# Identify potential roadblocks
proactive_intelligence["potential_issues"] = identify_potential_issues(
session_context, memory_insights
)
# Suggest optimizations
proactive_intelligence["optimization_opportunities"] = suggest_optimizations(
session_context, memory_insights
)
# Personalize recommendations
proactive_intelligence["personalized_suggestions"] = personalize_recommendations(
session_context, target_persona
)
return proactive_intelligence
Context Presentation Templates
Enhanced Context Briefing for Persona Activation
# 🧠 Memory-Enhanced Context Restoration for {Target Persona}
## 📍 Current Project State
**Project**: {project_name} | **Phase**: {current_phase} | **Duration**: {session_duration}
**Last Activity**: {last_persona} completed {last_task} {time_ago}
**Progress Status**: {completion_percentage}% through {current_epic}
## 🎯 Your Role Context
**Activation Reason**: {why_this_persona_now}
**Expected Contribution**: {anticipated_value_from_persona}
**Key Stakeholders**: {relevant_other_personas_and_user}
## 📚 Relevant Memory Intelligence
### Successful Patterns (from {similar_situations_count} similar cases)
- ✅ **{Success Pattern 1}**: Applied in {project_example} with {success_metric}
- ✅ **{Success Pattern 2}**: Used {usage_frequency} times with {average_outcome}
- ✅ **{Success Pattern 3}**: Proven effective for {context_specifics}
### Lessons Learned
- ⚠️ **Avoid**: {anti_pattern} (caused issues in {failure_count} cases)
- 🔧 **Best Practice**: {optimization_approach} (improved outcomes by {improvement_metric})
- 💡 **Insight**: {strategic_insight} (discovered from {learning_source})
## 🚀 Proactive Recommendations
### Immediate Actions
1. **{Priority Action 1}** - {rationale_with_memory_support}
2. **{Priority Action 2}** - {rationale_with_memory_support}
### Optimization Opportunities
- **{Optimization 1}**: {memory_based_suggestion}
- **{Optimization 2}**: {efficiency_improvement}
### Potential Issues to Watch
- **{Risk 1}**: {early_warning_signs} → **Prevention**: {mitigation_strategy}
- **{Risk 2}**: {indicators_to_monitor} → **Response**: {response_plan}
## 🎨 Personalization Insights
**Your Working Style**: {learned_preferences}
**Effective Approaches**: {what_works_well_for_user}
**Communication Preferences**: {optimal_interaction_style}
## ❓ Contextual Questions for Validation
Based on memory patterns, please confirm:
1. {context_validation_question_1}
2. {context_validation_question_2}
3. {preference_confirmation_question}
---
💬 **Memory Access**: Use `/recall {topic}` or ask "What do you remember about..."
🔍 **Deep Dive**: Use `/insights` for additional proactive intelligence
Lightweight Context Summary (for experienced users)
# ⚡ Quick Context for {Target Persona}
**Current**: {project_phase} | **Last**: {previous_activity}
**Memory Insights**: {key_pattern} proven in {success_cases} similar cases
**Recommended**: {next_action} based on {success_probability}% success rate
**Watch For**: {primary_risk} (early signs: {warning_indicators})
**Ready to proceed with {suggested_approach}?**
Context Restoration Intelligence
Pattern Recognition Engine
def recognize_context_patterns(session_context, memory_base):
pattern_analysis = {
"workflow_stage": classify_workflow_stage(session_context),
"success_probability": calculate_success_probability(session_context, memory_base),
"risk_assessment": assess_contextual_risks(session_context, memory_base),
"optimization_potential": identify_optimization_opportunities(session_context),
"user_alignment": assess_user_preference_alignment(session_context)
}
return pattern_analysis
def classify_workflow_stage(session_context):
stage_indicators = {
"project_initiation": ["no_prd", "analyst_activity", "brainstorming"],
"requirements_definition": ["prd_draft", "pm_activity", "scope_discussion"],
"architecture_design": ["architect_activity", "tech_decisions", "component_design"],
"development_preparation": ["po_activity", "story_creation", "validation"],
"active_development": ["dev_activity", "implementation", "testing"],
"refinement_cycle": ["multiple_persona_switches", "iterative_changes"]
}
return match_stage_indicators(session_context, stage_indicators)
Success Prediction Algorithm
def calculate_success_probability(current_context, memory_insights):
success_factors = {
"pattern_match_strength": calculate_pattern_similarity(current_context, memory_insights),
"context_completeness": assess_context_completeness(current_context),
"resource_availability": evaluate_resource_readiness(current_context),
"risk_mitigation": assess_risk_preparation(current_context, memory_insights),
"user_engagement": evaluate_user_engagement_patterns(current_context)
}
weighted_score = calculate_weighted_success_score(success_factors)
confidence_interval = calculate_confidence_interval(memory_insights.sample_size)
return {
"success_probability": weighted_score,
"confidence": confidence_interval,
"key_factors": identify_critical_success_factors(success_factors),
"improvement_opportunities": suggest_probability_improvements(success_factors)
}
Memory Creation During Context Restoration
Context Restoration Outcome Tracking
def track_context_restoration_effectiveness():
restoration_memory = {
"type": "context_restoration",
"session_context": current_session_state,
"memory_insights_provided": memory_intelligence_summary,
"persona_activation_success": measure_activation_effectiveness(),
"user_satisfaction": capture_user_feedback(),
"task_completion_improvement": measure_efficiency_gains(),
"accuracy_of_predictions": validate_proactive_insights(),
"learning_opportunities": identify_restoration_improvements()
}
add_memories(restoration_memory, tags=["context-restoration", "effectiveness", "learning"])
Proactive Intelligence Validation
def validate_proactive_insights(provided_insights, actual_outcomes):
validation_results = {}
for insight_type, predictions in provided_insights.items():
validation_results[insight_type] = {
"accuracy": calculate_prediction_accuracy(predictions, actual_outcomes),
"usefulness": measure_insight_application_rate(predictions),
"impact": assess_outcome_improvement(predictions, actual_outcomes)
}
# Update memory intelligence based on validation
update_proactive_intelligence_patterns(validation_results)
return validation_results
Integration with Persona Activation
Pre-Activation Context Assembly
def prepare_persona_activation_context(target_persona, session_state):
# 1. Load immediate session context
immediate_context = load_session_context()
# 2. Gather memory intelligence
memory_intelligence = gather_memory_intelligence(immediate_context, target_persona)
# 3. Generate proactive insights
proactive_insights = generate_proactive_insights(
immediate_context, memory_intelligence, target_persona
)
# 4. Synthesize comprehensive context
comprehensive_context = synthesize_context(
immediate_context, memory_intelligence, proactive_insights
)
# 5. Personalize for target persona
personalized_context = personalize_context(comprehensive_context, target_persona)
return personalized_context
Post-Activation Context Validation
def validate_context_restoration_success(persona_response, user_feedback):
validation_metrics = {
"context_completeness": assess_context_gaps(persona_response),
"memory_insight_relevance": evaluate_memory_application(persona_response),
"proactive_intelligence_value": measure_proactive_insight_usage(persona_response),
"user_satisfaction": capture_user_satisfaction(user_feedback),
"efficiency_improvement": measure_time_to_productivity(persona_response)
}
# Create learning memory for future context restoration improvement
create_context_restoration_learning_memory(validation_metrics)
return validation_metrics
Error Handling & Fallback Strategies
Memory System Unavailable
def fallback_context_restoration():
# Enhanced session state analysis
enhanced_session_context = analyze_session_state_deeply()
# Pattern recognition from session data
local_patterns = extract_patterns_from_session()
# Heuristic-based recommendations
heuristic_insights = generate_heuristic_insights(enhanced_session_context)
# Clear capability communication
communicate_reduced_capability_scope()
return create_fallback_context_briefing(
enhanced_session_context, local_patterns, heuristic_insights
)
Memory Query Failures
def handle_memory_query_failures(failed_queries, session_context):
# Attempt alternative query formulations
alternative_queries = reformulate_queries(failed_queries)
# Use cached memory insights if available
cached_insights = retrieve_cached_memory_insights(session_context)
# Generate context with available information
partial_context = create_partial_context(cached_insights, session_context)
# Flag limitations clearly
flag_context_limitations(failed_queries)
return partial_context
Quality Assurance & Continuous Improvement
Context Quality Metrics
- Relevance Score: How well memory insights match current context needs
- Completeness Score: Coverage of important contextual factors
- Accuracy Score: Correctness of proactive predictions and insights
- Usefulness Score: Practical value of context information for persona activation
- Efficiency Score: Time saved through effective context restoration
Continuous Learning Integration
def continuous_context_restoration_learning():
# Analyze recent context restoration outcomes
recent_restorations = get_recent_context_restorations()
# Identify improvement patterns
improvement_opportunities = analyze_restoration_effectiveness(recent_restorations)
# Update context restoration algorithms
update_context_intelligence(improvement_opportunities)
# Refine memory query strategies
optimize_memory_query_patterns(recent_restorations)
# Enhance proactive intelligence generation
improve_proactive_insight_algorithms(recent_restorations)