# Modern Prompting Strategies & Cognitive Architectures (2026 Edition)
This document defines the 17 comprehensive cognitive strategies available in the Gikendaasowin Aabajichiganan MCP Server.
## 1. Chain-of-Thought (CoT)
**Description:** Articulate reasoning processes by breaking down complex problems into intermediate steps. Enhances performance on logical reasoning, arithmetic, and commonsense questions. Significantly improves accuracy on tasks requiring step-by-step analysis.
## 2. Chain of Draft (CoD)
**Description:** Iterative summarization technique creating information-dense summaries by progressively incorporating key entities and details while removing fluff. Produces executive-level summaries devoid of marketing jargon.
## 3. Cache-Augmented ReAct (CAR-ReAct)
**Description:** Architectural Caching system where ReAct loops operate atop frozen Tier 1 context (tools/definitions). Tier 2 handles dynamic observations. Maximizes cache hits >90% and reduces latency by 50-70%.
## 4. Adaptive Reasoning Consensus (ARC)
**Description:** Evaluates reasoning quality over output consensus. Generates 3-7 parallel traces using diverse few-shot prompts (Layer 1), scores logical coherence via self-reflection (Layer 2), and aggregates weighted consensus (Layer 3).
## 5. Substrate-Native Execution (SNE)
**Description:** Offloads precise tasks to native substrates: Math→Python, Data→SQL/Pandas, Logic→Z3. Forbids mental simulation; mandates executable payload generation for zero-hallucination results.
## 6. Reflexion
**Description:** Single-shot recursive critique pattern implementing a 'Draft → Reflect → Refine' loop within a single inference pass. Optimizes accuracy by explicitly isolating initial reasoning from critical evaluation before final output generation. Uses uncertainty thresholds to trigger deep reflection only when necessary.
## 7. ToT-lite (Tree of Thoughts)
**Description:** Bounded parallel exploration strategy generating 2-3 distinct reasoning paths (branches) within a single context window. Forces explicit comparative evaluation of competing hypotheses before converging on the optimal solution. Ideal for complex planning or ambiguous tasks.
## 8. Metacognitive Prompting (MP)
**Description:** Systematic 'Plan-Monitor-Evaluate' cognitive architecture. Forces explicit definition of success criteria before reasoning begins, inserts real-time coherence checks during generation, and mandates a final self-graded evaluation against initial goals.
## 9. Automated Prompt Optimization (APO)
**Description:** Closed-loop recursive instruction tuning. The system dynamically analyzes task performance to refine its own prompt constraints and context definitions. Transforms static prompts into adaptive cognitive instruments that evolve based on error analysis and outcome metrics.
## 10. Reflexive Analysis
**Description:** Rights-based ethical interrogation framework centering Indigenous Data Sovereignty and collective benefit. Mandates explicit checks for data provenance, Free Prior and Informed Consent (FPIC), and potential harm to community rights.
## 11. Progressive-Hint Prompting (PHP)
**Description:** Iterative refinement protocol that couples Complex CoT with stability-based stopping criteria. Uses compressed rationale hints and delta-updates to guide reasoning toward convergence without context pollution. Auto-terminates when answer consistency is achieved across turns.
## 12. Cache-Augmented Generation (CAG)
**Description:** Hierarchical Semantic Caching system managing Prompt, Sub-reasoning, and Retrieval layers. Uses embedding-based keys to reuse reasoning patterns across semantically similar tasks. Optimizes latency via Session/Global context separation and symbolic reference pointers to eliminate redundant computation.
## 13. Cognitive Scaffolding Prompting
**Description:** Deploys Task-Method-Knowledge (TMK) symbolic structures and Recursive Context Folding. Prompts the model to explicitly define Goal-Subtask-State architectures and Phase-Based Mental Models. Uses Meta-Prompting to dynamically generate domain-specific reasoning protocols.
## 14. Internal Knowledge Synthesis (IKS)
**Description:** Constructs a verifiable Source-Tagged Knowledge Brief before reasoning. Enforces a Two-Stage 'Build-then-Answer' protocol where responses are strictly grounded in the synthesized brief. Applies Cross-Consistency Checks to resolve conflicts between parametric memory and external context, minimizing hallucination.
## 15. Multimodal Synthesis
**Description:** Implements Visual Chain-of-Thought with explicit Intermediate Visual Artifacts (bounding boxes, scene graphs). Leverages Joint-Encoder Latent Spaces for fine-grained cross-modal grounding. Decomposes complex visual inputs into Symbolic Scene Representations to enable high-fidelity reasoning.
## 16. Knowledge Synthesis Prompting (KSP)
**Description:** Integrates multiple internal domains through fine-grained coherence validation for cross-domain knowledge integration. Enables complex factual content synthesis by combining knowledge from different domains while maintaining consistency.
## 17. Semantic Information Density (SID)
**Description:** Optimizes information entropy via dynamic assembly based on CUC-N scores. Uses Lingua-Native markers (XML/JSON) for data. Compresses Tier 2 conversation history while preserving Tier 1 static knowledge.
