Advanced Patterns

Advanced patterns extend the 16 core patterns to the full set of 88, adding reasoning capabilities across specialized domains. They follow the same API as the core patterns — build_context(), execute(), generic_prompt() — and integrate seamlessly with the intelligence layer.

All open source

Every pattern is open source, ships in every install, and runs offline with your own LLM API key. No license keys, no tiers.

Importing Advanced Patterns

from mycontext.templates.enterprise.decision import DecisionFramework
from mycontext.templates.enterprise.systems import FeedbackLoopIdentifier
from mycontext.templates.enterprise.metacognition import MetacognitiveMonitor

# Same API as free patterns
result = DecisionFramework().execute(
    provider="openai",
    decision="Which cloud provider to standardize on",
    options=["AWS", "GCP", "Azure"],
    criteria=["cost", "reliability", "team expertise", "vendor lock-in risk"],
)

The Advanced Patterns

Specialized Intelligence (3 patterns)

Research-validated templates for RAG pipelines and memory — the most critical challenges in production AI systems.

Pattern	What it does	Key inputs	Research
QueryPlanner	Pre-retrieval query analysis: classify, decompose, rewrite (HyDE + step-back), and plan retrieval strategy. Use before RagAnswerer for best results.	query, task_type, domain	Validated
RagAnswerer	Grounded RAG generation with citation, abstention, and evidence extraction. Applies CRAG, Self-RAG, and Chain-of-Note principles. +15% evidence recall over plain RAG prompts.	question, context, mode	Validated
MemoryCompressor	Structured state extraction from conversations — entities, decisions, constraints, key numbers. 2x recall over progressive summarization at scale. Plugs into any framework's memory middleware.	content, intent, existing_memory, goal	Validated

from mycontext.templates.enterprise.specialized import QueryPlanner, RagAnswerer, MemoryCompressor

# Full RAG pipeline: plan query → retrieve → answer
planner = QueryPlanner()
plan = planner.execute(
    provider="openai",
    query="What are the main risks of our Q3 expansion plan?",
    task_type="multi-hop",
)

# Use the plan to guide retrieval, then answer with grounding
rag = RagAnswerer()
result = rag.execute(
    provider="openai",
    question=plan.response,
    context=retrieved_chunks,
    mode="answer",
)

# Memory compression for agent sessions
mc = MemoryCompressor()
result = mc.execute(
    provider="openai",
    content=conversation_history,
    intent="session",  # or "progressive" for updates
)

Advanced Analysis (4 patterns)

Deeper analytical capabilities beyond the free DataAnalyzer and QuestionAnalyzer.

Pattern	What it does
TrendIdentifier	Detect and extrapolate trend lines with statistical confidence intervals
GapAnalyzer	Systematic gap analysis between current state and desired state
SWOTAnalyzer	Full SWOT with strategic implications and prioritized actions
AnomalyDetector	Statistical anomaly detection with severity scoring and causal hypotheses

Advanced Reasoning (2 patterns)

Pattern	What it does
CausalReasoner	Build causal models with confounders, mechanisms, and intervention analysis
AnalogicalReasoner	Structured analogical transfer — map solutions from other domains

Advanced Creative (4 patterns)

Pattern	What it does
IdeaGenerator	SCAMPER + design thinking hybrid for product innovation
InnovationFramework	Jobs-to-be-Done framework for disruptive opportunity identification
DesignThinker	Full design thinking cycle: empathize → define → ideate → prototype → test
MetaphorGenerator	Construct precise explanatory metaphors for complex concepts

Advanced Communication (5 patterns)

Pattern	What it does
SimplificationEngine	Multi-level simplification with readability scoring
PersuasionFramework	Aristotle's ethos/pathos/logos persuasion architecture
NarrativeBuilder	Construct compelling narratives with arc, tension, and resolution
FeedbackComposer	Structured feedback using SBI (Situation-Behavior-Impact) model
ClarityOptimizer	Detect and resolve ambiguity, vagueness, and jargon systematically

Advanced Planning (3 patterns)

Pattern	What it does
ResourceAllocator	Optimal resource allocation with constraint modeling
PrioritySetter	Multi-criteria prioritization with RICE, ICE, and custom frameworks
DeadlineManager	Timeline analysis with critical path identification and buffer optimization

Advanced Specialized (5 patterns)

Pattern	What it does
ContentOutliner	Hierarchical content architecture with narrative flow
AmbiguityResolver	Systematically identify and resolve ambiguous requirements
RiskMitigator	Extends RiskAssessor with FMEA and bowtie analysis
ImpactAssessor	Multi-stakeholder impact assessment with quantified projections
ConceptExplainer	Layered explanations calibrated to expertise level (Feynman technique)

Decision Patterns (5 patterns)

Comprehensive decision support frameworks.

Pattern	What it does	Key inputs
DecisionFramework	Structured decision analysis with criteria weighting	decision, options, criteria
ComparativeAnalyzer	Head-to-head comparison with normalized scoring	items, dimensions
TradeoffAnalyzer	Explicit tradeoff mapping across multiple dimensions	options, tradeoffs
MultiObjectiveOptimizer	Pareto-optimal solution finding with constraint satisfaction	objectives, constraints
CostBenefitAnalyzer	Quantified CBA with sensitivity analysis and NPV	costs, benefits, timeframe

from mycontext.templates.enterprise.decision import DecisionFramework

result = DecisionFramework().execute(
    provider="openai",
    decision="Choose our primary database technology",
    options=["PostgreSQL", "MongoDB", "DynamoDB"],
    criteria={
        "query_flexibility": 0.30,
        "operational_complexity": 0.25,
        "cost": 0.20,
        "team_expertise": 0.15,
        "scalability": 0.10,
    },
)

Problem Solving (6 patterns)

Pattern	What it does
ProblemDecomposer	MECE decomposition into independently solvable sub-problems
BottleneckIdentifier	Throughput analysis using Theory of Constraints methodology
ConstraintOptimizer	Optimize outcomes within a constraint set
DependencyMapper	Map dependency graphs with critical path analysis
EfficiencyAnalyzer	Waste identification and process optimization (Lean principles)
TradeSpaceExplorer	Explore the full solution space before committing to one option

Temporal Reasoning (3 patterns)

Pattern	What it does
TemporalSequenceAnalyzer	Analyze causation and dependencies across time
FutureScenarioPlanner	Extended ScenarioPlanner with morphological analysis
HistoricalContextMapper	Map historical precedents and their implications for current decisions

Diagnostic (3 patterns)

Medical-grade diagnostic reasoning methodology applied to any domain.

Pattern	What it does
DiagnosticRootCauseAnalyzer	FMEA-integrated root cause analysis (extends free RootCauseAnalyzer)
DifferentialDiagnoser	Differential diagnosis — systematic elimination of alternative explanations
SystemHealthAuditor	Full-system health audit across multiple diagnostic dimensions

Synthesis (3 patterns)

Pattern	What it does
HolisticIntegrator	Cross-domain synthesis with emergence detection
PatternRecognitionEngine	Identify recurring patterns in complex, noisy data
CrossDomainSynthesizer	Transfer insights across domain boundaries

Systems Thinking (6 patterns)

The most sophisticated pattern category — models complex adaptive systems.

Pattern	What it does
FeedbackLoopIdentifier	Identify reinforcing and balancing feedback loops
LeveragePointFinder	Find highest-impact intervention points in a system (Meadows' leverage points)
EmergenceDetector	Identify emergent properties and non-linear dynamics
SystemArchetypeAnalyzer	Match system behavior to known archetypes (fixes that fail, tragedy of the commons, etc.)
CausalLoopDiagrammer	Build formal causal loop diagrams with polarity and delay notation
StockFlowAnalyzer	Model stocks, flows, and accumulation dynamics

from mycontext.templates.enterprise.systems import LeveragePointFinder

result = LeveragePointFinder().execute(
    provider="openai",
    system="Enterprise SaaS customer retention system",
    goal="Identify where small changes would have the largest impact on churn",
    context="Current churn: 3.2%/month. Top churn drivers: feature gaps, support responsiveness, pricing",
)

Metacognition (5 patterns)

Patterns that think about thinking — improve reasoning quality and decision processes.

Pattern	What it does
MetacognitiveMonitor	Monitor and improve the quality of reasoning processes
SelfRegulationFramework	Structure self-correction and adaptive thinking
CognitiveStrategySelector	Select the optimal cognitive strategy for a given problem type
LearningFromExperience	Structured after-action review and lessons-learned extraction
ErrorDetectionFramework	Systematic identification of cognitive biases and reasoning errors

Ethical Reasoning (5 patterns)

Structured application of ethical frameworks to decisions and situations.

Pattern	What it does
EthicalFrameworkAnalyzer	Apply multiple ethical frameworks (consequentialist, deontological, virtue) to a situation
MoralDilemmaResolver	Structured approach to genuine moral dilemmas without easy answers
StakeholderEthicsAssessor	Assess ethical implications across all affected stakeholders
ValueConflictNavigator	Navigate situations where values genuinely conflict
ConsequentialistAnalyzer	Rigorous consequence mapping with probability-weighted impact

Learning (5 patterns)

Educational and knowledge-building patterns.

Pattern	What it does
ScaffoldingFramework	Build knowledge from existing foundations progressively
SpacedRepetitionOptimizer	Optimize learning sequences for retention
ZoneOfProximalDevelopment	Calibrate challenge level to maximize learning
CognitiveLoadManager	Reduce extraneous cognitive load in explanations
ConceptualChangeAnalyzer	Identify and address conceptual misconceptions

Evaluation (5 patterns)

Assessment and feedback patterns for systematic quality evaluation.

Pattern	What it does
RubricDesigner	Create evaluation rubrics with clear criteria and descriptors
FormativeAssessmentFramework	Ongoing assessment to guide improvement
SummativeEvaluator	Final evaluation against defined criteria
PeerAssessmentStructure	Structure effective peer review processes
SelfAssessmentGuide	Guide structured self-evaluation

Pattern Combinations

Advanced patterns are designed to chain with each other and with the core patterns:

from mycontext.templates.enterprise.decision import TradeoffAnalyzer
from mycontext.templates.enterprise.systems import LeveragePointFinder
from mycontext.templates.free.planning import ScenarioPlanner

# Step 1: Map tradeoffs in the decision space
tradeoffs = TradeoffAnalyzer().execute(
    provider="openai",
    options=["Build in-house", "Use third-party API", "Acquire startup"],
    tradeoffs=["control", "cost", "speed", "technical risk"],
)

# Step 2: Find the highest-leverage choice
leverage = LeveragePointFinder().execute(
    provider="openai",
    system="AI feature development ecosystem",
    context=tradeoffs.response[:1000],
)

# Step 3: Plan for uncertainty
scenarios = ScenarioPlanner().execute(
    provider="openai",
    topic="Build vs. buy decision for core AI feature",
    timeframe="18 months",
)

License

All 88 cognitive patterns — core and advanced — are open source under the MIT license and ship in every pip install mycontext-ai. There are no tiers and no license keys. Run any pattern offline with your own LLM API key.