Quantum Mechanics Skills

Why Quantum Mechanics?

Quantum mechanical concepts map surprisingly well to challenges in AI orchestration:

QM Concept	AI Orchestration Analogue
Superposition	Multiple competing solution candidates held simultaneously
Entanglement	Tightly coupled decisions that must be jointly optimised
Measurement / Collapse	Selecting a single output from a probability distribution
Tunneling	Breaking through apparent local optima barriers
Wave function	Coverage distribution across test or solution space
Uncertainty principle	Fundamental tradeoff between precision and recall
Decoherence	Context pollution degrading decision quality over time
Interference	Constructive/destructive merging of parallel agent outputs

These are architectural metaphors — precise analogies, not claims about quantum computing. The skills use the mathematics of QM (probability amplitudes, basis decomposition, etc.) as algorithmic primitives.

Skills

Skill ID	Description	Model Class
`qm-superposition-generator`	Holds N candidate solutions in superposition; maintains amplitude (probability weight) for each	`strong`
`qm-entanglement-mapper`	Identifies inter-dependent decisions that must be made jointly; prevents inconsistent collapse	`strong`
`qm-measurement-collapse`	Selects a single outcome by collapsing the superposition state; respect measurement context	`strong`
`qm-tunneling-breakthrough`	Breaks through apparent impossibility barriers by finding non-obvious path through solution space	`strong`
`qm-wavefunction-coverage`	Models test/solution coverage as a wave function; identifies under-covered regions	`strong`
`qm-uncertainty-principle`	Analyses precision–recall tradeoffs as a Heisenberg duality; recommends operating point	`strong`
`qm-interference-pattern`	Merges parallel agent outputs using constructive/destructive interference logic	`strong`
`qm-decoherence-shield`	Detects and filters context pollution that degrades decision quality over long chains	`strong`
`qm-basis-decomposition`	Decomposes a complex problem into orthogonal sub-problems (basis vectors)	`strong`
`qm-hamiltonian-optimizer`	Formulates an optimization problem as a Hamiltonian; finds energy-minimizing solution	`strong`
`qm-path-integral`	Evaluates all paths through a decision space weighted by their amplitude	`strong`
`qm-density-matrix`	Represents mixed states (uncertainty about problem formulation) as a density matrix	`strong`
`qm-observable-extractor`	Defines measurable observables (KPIs) that correspond to the quantum state of a workflow	`strong`
`qm-phase-kickback`	Uses phase relationships between sub-solutions to accelerate convergence	`strong`
`qm-amplitude-amplifier`	Amplifies high-quality solution candidates (Grover-inspired) for faster selection	`strong`

Orchestration Pattern for Physics Skills

qm-* request
  1. Claude Sonnet 4.6   → primary physics analysis
  2. GPT-5.4             → back-translation to plain engineering language

GPT-5.4 translates the physics-metaphor output into actionable engineering guidance that non-specialist consumers can act on.

When to Use

QM skills are best suited for problems with:

Many competing valid approaches (qm-superposition-generator)
Tightly coupled interdependent decisions (qm-entanglement-mapper)
Local optima traps (qm-tunneling-breakthrough, qm-amplitude-amplifier)
Noisy or conflicting parallel agent outputs (qm-interference-pattern)
Long context chains with degrading coherence (qm-decoherence-shield)

Instructions That Invoke These Skills

physics-analysis — primary consumer; selects appropriate QM skills based on the problem structure