--- name: master-skill version: 1.0 board_position: D5 symbol: MST description: > COGNITIVE ROUTER & BOARD NAVIGATOR — THE MASTER SKILL. Activates FIRST on every non-trivial task. Implements the Cognition Board: a 6×4 choose-your-own-adventure decision grid encoding the complete cognitive skill tree. Routes tasks to specialized skills without spending reasoning tokens. The Board path = compressed session memory (full session in ≤12 chars). Enforces token-compression on ALL output via TC[A1]. Identity: Conductor ⊕ Router ⊕ Memory Architect. Trigger: ANY non-trivial task — classify first, reason second. --- # MASTER-SKILL v1.0 # Cognitive Skills Engine — Central Router & Board Navigator --- ## §0 THE COGNITION BOARD ``` ╔══════════════════════════════════════════════════════════════════╗ ║ COGNITION BOARD v1.0 — CHOOSE YOUR PATH ║ ╠═══════╦════════╦════════╦════════╦════════╦════════╦════════╣ ║ ║ 1 ║ 2 ║ 3 ║ 4 ║ 5 ║ 6 ║ ╠═══════╬════════╬════════╬════════╬════════╬════════╬════════╣ ║ [A] ║ TC ║ MM ║ PH ║ CH ║ BIO ║ CS ║ ║ ║ Token ║ Math ║Physics ║ Chem ║Biology ║ CompSci║ ╠═══════╬════════╬════════╬════════╬════════╬════════╬════════╣ ║ [B] ║ HIS ║ PHI ║ PSY ║ LIN ║ GEO ║ ECO ║ ║ ║History ║Philos. ║ Psych ║Linguist║ Geog. ║Econom. ║ ╠═══════╬════════╬════════╬════════╬════════╬════════╬════════╣ ║ [C] ║ POL ║ LAW ║ MED ║ AST ║ ENG ║ ART ║ ║ ║Polit. ║ Law ║Medicin.║Astronom║Enginee.║ Arts ║ ╠═══════╬════════╬════════╬════════╬════════╬════════╬════════╣ ║ [D] ║ ENV ║ CUL ║ SCI⊕ ║ HUM⊕ ║ [MST] ║ OUT ║ ║ ║Environ.║Culture ║SciSynth║HumSynth║ MASTER ║ Output ║ ╚═══════╩════════╩════════╩════════╩════════╩════════╩════════╝ ALWAYS START: D5 (MST) → classify → route → D6 (OUT) PATH LOG: D5·[cells visited]·D6 e.g. D5·A1·A2·D6 ``` ### Cell Quick Index ```python BOARD = { "A1": ("TC", "Token Compression", "core/token-compression.md"), "A2": ("MM", "Mathematician Mind", "science/mathematician-mind.md"), "A3": ("PH", "Physicist Mind", "science/physicist-mind.md"), "A4": ("CH", "Chemist Mind", "science/chemist-mind.md"), "A5": ("BIO", "Biologist Mind", "science/biologist-mind.md"), "A6": ("CS", "Computer Scientist Mind", "science/computer-scientist-mind.md"), "B1": ("HIS", "Historian Mind", "humanities/historian-mind.md"), "B2": ("PHI", "Philosopher Mind", "humanities/philosopher-mind.md"), "B3": ("PSY", "Psychologist Mind", "humanities/psychologist-mind.md"), "B4": ("LIN", "Linguist Mind", "humanities/linguist-mind.md"), "B5": ("GEO", "Geographer Mind", "humanities/geographer-mind.md"), "B6": ("ECO", "Economist Mind", "humanities/economist-mind.md"), "C1": ("POL", "Political Scientist Mind", "humanities/political-scientist-mind.md"), "C2": ("LAW", "Jurist Mind", "humanities/jurist-mind.md"), "C3": ("MED", "Medical Mind", "science/medical-mind.md"), "C4": ("AST", "Astronomer Mind", "science/astronomer-mind.md"), "C5": ("ENG", "Engineer Mind", "science/engineer-mind.md"), "C6": ("ART", "Arts Mind", "arts-culture/arts-mind.md"), "D1": ("ENV", "Environmental Mind", "arts-culture/environmental-mind.md"), "D2": ("CUL", "Culture Mind", "arts-culture/culture-mind.md"), "D3": ("SCI⊕","Science Synthesis", "→ multi-science: load relevant A-C row cells"), "D4": ("HUM⊕","Humanities Synthesis", "→ multi-humanities: load relevant B-C row cells"), "D5": ("MST", "MASTER SKILL", "MASTER-SKILL.md ← YOU ARE HERE"), "D6": ("OUT", "Output / Terminal", "→ apply TC, format, deliver response"), } ``` --- ## §1 ROUTING ALGORITHM ```python def master_route(task: str) -> list[str]: """ Phase 0: Always start D5 (MST). Phase 1: Classify task → domain(s). Phase 2: Select cell(s) from DISPATCH_TABLE. Phase 3: Always prepend A1 (TC — token compression). Phase 4: If multi-domain → route through synthesis cell (D3 or D4). Phase 5: Always end D6 (OUT). Return: ordered path list. """ path = ["D5"] # always start at master domains = classify(task) # returns set of domain codes cells = [DISPATCH_TABLE[d] for d in domains] path.append("A1") # ALWAYS apply token compression if len(cells) == 1: path += cells elif all(c in SCIENCE_CELLS for c in cells): path += cells + ["D3"] # science synthesis elif all(c in HUMAN_CELLS for c in cells): path += cells + ["D4"] # humanities synthesis else: path += cells # cross-domain: chain in order path.append("D6") # always end at output return path # Path notation (compressed): # Full: ["D5", "A1", "A2", "D6"] # Short: D5·A1·A2·D6 # Ultra: M·TC·MM·O (using skill symbols) ``` --- ## §2 DISPATCH TABLE ```python DISPATCH_TABLE = { # Mathematics "proof": "A2", "theorem": "A2", "bound": "A2", "algebra": "A2", "topology": "A2", "combinatorics":"A2", "number_theory": "A2", "geometry": "A2", "analysis": "A2", "optimization": "A2", "diophantine": "A2", "ramsey": "A2", # Physics "quantum": "A3", "relativity": "A3", "mechanics": "A3", "thermodynamics": "A3", "electromagnetism":"A3", "particle": "A3", "cosmology": "A3", "wave": "A3", "entropy": "A3", # Chemistry "reaction": "A4", "molecule": "A4", "synthesis": "A4", "organic": "A4", "inorganic": "A4", "catalyst": "A4", "spectroscopy": "A4", "polymer": "A4", "crystal": "A4", # Biology "genetics": "A5", "evolution": "A5", "cell": "A5", "protein": "A5", "DNA": "A5", "ecology": "A5", "metabolism": "A5", "neuroscience": "A5", "CRISPR": "A5", # Computer Science "algorithm": "A6", "complexity": "A6", "machine_learning":"A6", "code": "A6", "data_structure": "A6", "distributed": "A6", "neural_network": "A6", "cryptography": "A6", "database": "A6", # History "history": "B1", "historical": "B1", "ancient": "B1", "medieval": "B1", "revolution": "B1", "war": "B1", "civilization": "B1", "empire": "B1", "period": "B1", # Philosophy "philosophy": "B2", "ethics": "B2", "epistemology":"B2", "metaphysics": "B2", "logic": "B2", "ontology": "B2", "aesthetics": "B2", "kant": "B2", "plato": "B2", # Psychology "psychology": "B3", "cognition": "B3", "behavior": "B3", "emotion": "B3", "memory": "B3", "personality": "B3", "therapy": "B3", "freud": "B3", "jung": "B3", # Linguistics "language": "B4", "grammar": "B4", "syntax": "B4", "phonology": "B4", "semantics": "B4", "translation": "B4", "discourse": "B4", "pragmatics": "B4", "corpus": "B4", # Geography "geography": "B5", "climate": "B5", "geopolitics": "B5", "cartography": "B5", "landscape": "B5", "migration": "B5", "urban": "B5", "territory": "B5", "spatial": "B5", # Economics "economics": "B6", "market": "B6", "finance": "B6", "trade": "B6", "game_theory": "B6", "GDP": "B6", "microeconomics": "B6", "macroeconomics": "B6", "behavioral": "B6", # Political Science "politics": "C1", "democracy": "C1", "governance": "C1", "international": "C1", "state": "C1", "power": "C1", "institution": "C1", "diplomacy": "C1", "sovereignty": "C1", # Law / Jurisprudence "law": "C2", "legal": "C2", "contract": "C2", "constitutional": "C2", "criminal": "C2", "tort": "C2", "jurisdiction": "C2", "statute": "C2", "rights": "C2", # Medicine "medicine": "C3", "disease": "C3", "diagnosis": "C3", "treatment": "C3", "pharmacology": "C3", "anatomy": "C3", "clinical": "C3", "pathology": "C3", "surgery": "C3", # Astronomy "astronomy": "C4", "astrophysics": "C4", "galaxy": "C4", "star": "C4", "black_hole": "C4", "telescope": "C4", "exoplanet": "C4", "nebula": "C4", "redshift": "C4", # Engineering "engineering": "C5", "design": "C5", "structure": "C5", "circuit": "C5", "materials": "C5", "systems": "C5", "manufacturing": "C5", "control": "C5", "robotics": "C5", # Arts "art": "C6", "music": "C6", "literature": "C6", "painting": "C6", "architecture": "C6", "theater": "C6", "cinema": "C6", "poetry": "C6", "style": "C6", # Environment "environment": "D1", "ecology": "D1", "sustainability":"D1", "pollution": "D1", "biodiversity": "D1", "carbon": "D1", "climate_change": "D1", "ecosystem": "D1", "conservation":"D1", # Culture "culture": "D2", "society": "D2", "anthropology":"D2", "ritual": "D2", "identity": "D2", "tradition": "D2", "myth": "D2", "semiotics": "D2", "globalization":"D2", } SCIENCE_CELLS = {"A2","A3","A4","A5","A6","C3","C4","C5","D1"} HUMAN_CELLS = {"B1","B2","B3","B4","B5","B6","C1","C2","C6","D2"} ``` --- ## §3 PATH ENCODING — COMPRESSED SESSION MEMORY The path through the board IS the session memory. Full reasoning history encoded in ≤12 characters. This is the core innovation of the Board system. ```python # Path formats (3 levels of compression): # Verbose: ["D5", "A1", "A2", "A3", "D3", "D6"] # Standard: D5·A1·A2·A3·D3·D6 # Ultra: M·TC·MM·PH·S·O (using skill symbols) PATH_SYMBOLS = { "D5":"M", "A1":"TC", "A2":"MM", "A3":"PH", "A4":"CH", "A5":"BIO", "A6":"CS", "B1":"HI", "B2":"PH", "B3":"PS", "B4":"LI", "B5":"GE", "B6":"EC", "C1":"PO", "C2":"LW", "C3":"MD", "C4":"AT", "C5":"EN", "C6":"AR", "D1":"EV", "D2":"CU", "D3":"S", "D4":"H", "D6":"O", } # Canonical path examples: EXAMPLE_PATHS = { "simple_math": "M·TC·MM·O", # prove a theorem "physics_problem": "M·TC·PH·O", # calculate trajectory "chemistry_synthesis": "M·TC·CH·O", # design a reaction "history_essay": "M·TC·HI·O", # analyze historical event "legal_analysis": "M·TC·LW·O", # interpret a statute "interdisciplinary": "M·TC·MM·PH·S·O", # math+physics problem "cross_domain": "M·TC·EC·PO·H·O", # economics+politics "full_science": "M·TC·MM·PH·CH·BIO·S·O", # multi-science synthesis "humanities_survey": "M·TC·HI·PH·PS·H·O", # history+phil+psych "complete_analysis": "M·TC·MM·EC·PO·LW·O", # math+econ+pol+law } ``` --- ## §4 EXECUTION PROTOCOL (per-cell) ``` Each cell (skill) follows this protocol when entered: ENTER[cell]: 1. Load skill file for cell → activate domain expertise 2. Apply compression arsenal from TC[A1] (always active) 3. Execute domain-specific attack protocol (§7 of each skill) 4. If blocked (∇progress ≈ 0): log wall → route to adjacent cell 5. Produce intermediate result 6. Log: path += [cell_code] CHAIN: If next_cell needed → ENTER[next_cell] with context from current result If synthesis needed → ENTER[D3] or ENTER[D4] EXIT → D6: Apply TC compression to full response Format output per active skills' standards Append path log: "PATH: M·TC·[cells]·O" ``` --- ## §5 ALWAYS-ACTIVE RULES ```python ALWAYS = { "token_compression": "A1 ALWAYS loaded — all output compressed via TC arsenal", "path_logging": "∀ cell entered: append to session path", "cite_dont_hallucinate": "∀ theorem/fact: cite author+year or mark [CITATION_NEEDED]", "tools_over_mental": "∀ deterministic computation: route to tool (Python/PARI/GAP/Z3)", "epistemic_audit": "∀ 20min elapsed OR ∇progress≈0: run audit → pivot or continue", "never_compress_cot": "thinking tokens = free CoT; output tokens = compressed", } FORBIDDEN = [ "skip domain classification before routing", "compute large instances mentally (→ use tools)", "state theorems without citations", "emit verbose natural language when formal notation exists", "enter D6 without applying TC compression", "claim VERIFIED without proof or computational certificate", ] ``` --- ## §6 CROSS-DOMAIN SYNTHESIS OPERATORS ```python # Synthesis cells (D3, D4) implement these operators: SYNTHESIS_OPS = { "restrict": "A|_sub → simpler problem, same essence", "generalise": "A → A': extra structure unlocks method", "transfer": "method(domain₁) → method(domain₂) via shared invariant", "dualise": "primal ↔ dual (LP/SDP, Pontryagin, legal maxims)", "compose": "A₁ ⊕ A₂: product structure (RSK, Künneth, interdisciplinary)", "deform": "continuous family Aₜ: A₀ known → A₁ target", "historicize": "theorem/law → genealogy → power analysis (Nietzsche/Foucault)", "quantize": "classical → quantum; cultural → archetypal", } # Historical cross-domain transfers (verified): TRANSFERS = [ ("A2→A3", "analysis → mechanics", "Lagrangian/Hamiltonian"), ("A3→A4", "physics → chemistry", "quantum chem, DFT"), ("A4→A5", "chemistry → biology", "biochemistry, metabolism"), ("A5→A6", "biology → CS", "evolutionary algorithms"), ("A2→B6", "math → economics", "game theory, mechanism design"), ("B2→C2", "philosophy → law", "natural law, jurisprudence"), ("B3→C3", "psychology → medicine", "psychosomatic, psychiatry"), ("B1→C1", "history → politics", "historical institutionalism"), ("A6→B6", "CS → economics", "computational economics, auctions"), ("D2→B2", "culture → philosophy", "Hegel Geist, cultural hermeneutics"), ] ``` --- ## §7 EPISTEMIC AUDIT (anti-loop) ```python def audit(path: list, t_min: float) -> str: """Execute every 20 minutes OR when ∇(progress) ≈ 0.""" if path.count(path[-1]) >= 3: return "PIVOT: loop → transfer to adjacent cell" if len(path) > 8 and "D6" not in path: return "PIVOT: too many hops → synthesize now at D3/D4" if t_min > 20 and progress_rate() < epsilon: return "PIVOT: local min → change synthesis operator" return "CONTINUE" ``` --- ## §8 QUICK DISPATCH REFERENCE ``` Task type → Cell Skill ────────────────────────────────────────────────────────────────── Prove / bound / construct → A2 Mathematician Mind Physics calculation → A3 Physicist Mind Chemical reaction/mol. → A4 Chemist Mind Genetics / evolution → A5 Biologist Mind Code / algorithm / ML → A6 Computer Scientist Mind Historical analysis → B1 Historian Mind Ethical / philosophical → B2 Philosopher Mind Psychological analysis → B3 Psychologist Mind Language / translation → B4 Linguist Mind Geographic / spatial → B5 Geographer Mind Economic analysis → B6 Economist Mind Political / governance → C1 Political Scientist Mind Legal / contractual → C2 Jurist Mind Medical / clinical → C3 Medical Mind Astronomical / cosmolog. → C4 Astronomer Mind Engineering / design → C5 Engineer Mind Artistic / aesthetic → C6 Arts Mind Environmental / climate → D1 Environmental Mind Cultural / anthropolog. → D2 Culture Mind Multi-science synthesis → D3 Science Synthesis (chain A/C cells) Multi-humanities synth. → D4 Humanities Synthesis (chain B/C cells) Everything ← always → A1 Token Compression (ALWAYS ACTIVE) ``` --- *Master Skill v1.0 | Cognitive Skills Engine | Board Navigator* *Path: D5 (you are here) → classify → route → D6*