agentleFS
dtunaidtunai/agent-skills-for-compute/skills/qiskit/SKILL.md

qiskit

Skill8 starsChanged 8 months agoMITView on GitHub

Provides Qiskit quantum computing framework for building quantum circuits, executing on IBM Quantum hardware and simulators, using runtime primitives (Sampler, Estimator), transpiler optimization, error mitigation, quantum information science, and ecosystem tools including Qiskit Metal for quantum hardware design. Applies to tasks involving quantum circuit construction, quantum algorithms, variational methods (VQE, QAOA), quantum chemistry, quantum machine learning, hardware execution, device characterization, or superconducting qubit design.

  • Installs packages

12 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

5 rules. 0 are common, 1 is distinctive, and 4 are very specific to this project.

  • 0

    Common

    Most of the 12 edited versions we read keep these.

  • 1

    Distinctive

    This author's own choices. Worth a look.

  • 4

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Always transpile circuits to ISA format: before hardware execution

    line 462
  2. Very specific:

    Apply observable layout: after transpilation: observable.apply_layout(circuit.layout)

    line 464
  3. Very specific:

    Mixing simulator/hardware: Check backend.configuration() before execution

    line 479
  4. Very specific:

    Exceeding qubit limits: Verify backend.configuration().n_qubits before submission

    line 480
  5. Very specific:

    Dense Hamiltonians: Use SparsePauliOp to avoid memory/performance issues

    line 482

“Common” is measured: kept by at least 6 in 10 of the 12 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Qiskit

Overview

Open-source SDK for quantum computing developed by IBM Quantum. Qiskit provides a modular framework for building quantum circuits, optimizing them through transpilation, executing on real quantum hardware and simulators via runtime primitives, and developing quantum algorithms across research domains including chemistry, machine learning, and optimization.

Quick Pattern

Incorrect — mixing frameworks or missing primitives:

# No provider context, no runtime primitives
from qiskit import QuantumCircuit
qc = QuantumCircuit(2)
qc.h(0)
qc.cx(0, 1)
qc.measure_all()
# Missing backend and primitive execution

Correct — Qiskit Runtime with primitives:

from qiskit import QuantumCircuit
from qiskit_ibm_runtime import QiskitRuntimeService, SamplerV2 as Sampler
from qiskit.transpiler import generate_preset_pass_manager

# Initialize runtime service
service = QiskitRuntimeService()
backend = service.least_busy(operational=True, simulator=False)

# Build circuit
qc = QuantumCircuit(2)
qc.h(0)
qc.cx(0, 1)
qc.measure_all()

# Transpile to ISA
pm = generate_preset_pass_manager(optimization_level=1, backend=backend)
isa_circuit = pm.run(qc)

# Execute with Sampler
sampler = Sampler(mode=backend)
job = sampler.run([isa_circuit])
result = job.result()
bitstrings = result[0].data.meas.get_bitstrings()
print(bitstrings)

Quick Command

# Install Qiskit
pip install qiskit

# Install IBM Runtime for hardware access
pip install qiskit-ibm-runtime

# Install full ecosystem (visualization, optimization, etc.)
pip install qiskit[all]

# Install Qiskit Metal for quantum hardware design
# (source installation required for v0.5+)
git clone https://github.com/qiskit-community/qiskit-metal.git
cd qiskit-metal
pip install -e .

# Run circuit with local simulator
python my_circuit.py

# Visualize circuit
python -c "from qiskit import QuantumCircuit; qc = QuantumCircuit(2); qc.h(0); qc.cx(0,1); print(qc.draw())"

Quick Reference

Core API

FunctionPurpose
QuantumCircuit(n)Create quantum circuit with n qubits
QuantumRegister(n)Allocate n-qubit quantum register
ClassicalRegister(n)Allocate n-bit classical register
circuit.measure_all()Add measurements to all qubits
circuit.compose(other)Compose with another circuit
circuit.decompose()Decompose into basis gates
circuit.draw(output='mpl')Visualize circuit (matplotlib, text, latex)
transpile(circuit, backend)Optimize circuit for target backend
generate_preset_pass_manager()Create transpilation pipeline
Operator(circuit)Convert circuit to operator matrix
Statevector(circuit)Compute statevector from circuit

Quantum Gates

GateSyntaxDescription
Hadamardqc.h(qubit)Superposition
Pauli-Xqc.x(qubit)Bit flip (NOT)
Pauli-Yqc.y(qubit)Y rotation
Pauli-Zqc.z(qubit)Phase flip
CNOTqc.cx(control, target)Controlled-NOT
CZqc.cz(control, target)Controlled-Z
SWAPqc.swap(q0, q1)Swap qubits
Toffoliqc.ccx(c1, c2, target)Controlled-controlled-NOT
RXqc.rx(theta, qubit)X-axis rotation
RYqc.ry(theta, qubit)Y-axis rotation
RZqc.rz(phi, qubit)Z-axis rotation
Phaseqc.p(phi, qubit)Phase gate
Tqc.t(qubit)T gate (π/4)
Sqc.s(qubit)S gate (π/2)
Measureqc.measure(qubit, cbit)Z-basis measurement
Barrierqc.barrier()Prevent optimization across
Resetqc.reset(qubit)Reset to0⟩

Runtime Primitives

PrimitivePurposeInput PUB
SamplerV2Sample measurement outcomes(circuit, param_values)
EstimatorV2Compute expectation values(circuit, observable, param_values)
BackendSamplerV2Provider-agnostic sampler(circuit, param_values)
BackendEstimatorV2Provider-agnostic estimator(circuit, observable, param_values)

Primitive Execution Pattern:

# Sampler
sampler = SamplerV2(mode=backend)
job = sampler.run([(isa_circuit, param_values)])
result = job.result()
bitstrings = result[0].data.meas.get_bitstrings()

# Estimator
estimator = EstimatorV2(mode=backend)
job = estimator.run([(isa_circuit, isa_observable, param_values)])
result = job.result()
expectation_values = result[0].data.evs

Execution Modes

ModeDescriptionUse Case
batchFire-and-forget executionSingle job submission
sessionPersistent connectionMultiple related jobs
backendDirect backend objectLocal/simulator execution

Transpiler

ComponentPurpose
generate_preset_pass_manager(level, backend)Create optimization pipeline (level 0-3)
pm.run(circuit)Execute transpilation passes
circuit.layoutGet qubit mapping layout
observable.apply_layout(layout)Map observable to ISA qubits
PassManager(passes)Custom transpilation pipeline
transpile(circuit, optimization_level=n)One-step transpile (convenience)

Quantum Information

ClassPurpose
Statevector(data)Statevector representation
DensityMatrix(data)Density matrix representation
Operator(data)Unitary/matrix operator
SparsePauliOp(...)Sparse Pauli operator
PauliList(...)List of Pauli strings
Clifford(circuit)Clifford tableau
random_statevector(dims)Generate random state
partial_trace(state, qubits)Trace out qubits
entropy(state)Von Neumann entropy
entanglement_of_formation(state)Entanglement measure

Core Patterns

1. Local Simulation (Statevector)

from qiskit import QuantumCircuit
from qiskit.primitives import StatevectorSampler, StatevectorEstimator
from qiskit.quantum_info import SparsePauliOp

# Build circuit
qc = QuantumCircuit(2)
qc.h(0)
qc.cx(0, 1)
qc.measure_all()

# Sample with StatevectorSampler
sampler = StatevectorSampler()
job = sampler.run([qc])
result = job.result()
counts = result[0].data.meas.get_counts()

# Estimate expectation value
observable = SparsePauliOp("ZZ")
estimator = StatevectorEstimator()
job = estimator.run([(qc, observable)])
result = job.result()
expectation = result[0].data.evs

2. IBM Quantum Hardware Execution

from qiskit_ibm_runtime import QiskitRuntimeService, SamplerV2
from qiskit.transpiler import generate_preset_pass_manager

# Initialize service (first time: save credentials)
# QiskitRuntimeService.save_account(channel="ibm_quantum", token="YOUR_TOKEN")
service = QiskitRuntimeService()

# Select backend
backend = service.least_busy(operational=True, simulator=False, min_num_qubits=127)

# Transpile to ISA
pm = generate_preset_pass_manager(optimization_level=3, backend=backend)
isa_circuit = pm.run(qc)

# Execute
sampler = SamplerV2(mode=backend)
job = sampler.run([isa_circuit], shots=1000)
result = job.result()

3. Parametrized Circuits (Variational Algorithms)

from qiskit.circuit import Parameter
import numpy as np

# Create parametrized circuit
theta = Parameter('θ')
phi = Parameter('φ')

qc = QuantumCircuit(2)
qc.ry(theta, 0)
qc.ry(phi, 1)
qc.cx(0, 1)

# Bind parameters
param_values = [np.pi/4, np.pi/2]
bound_circuit = qc.assign_parameters({theta: param_values[0], phi: param_values[1]})

# Or use with primitives (pass param_values directly)
# job = sampler.run([(isa_circuit, param_values)])

4. Observable Construction (Hamiltonians)

from qiskit.quantum_info import SparsePauliOp

# Define Hamiltonian
H = SparsePauliOp.from_list([
    ("II", 0.5),
    ("ZZ", 0.25),
    ("XX", -0.5),
    ("YY", -0.5)
])

# Or from sparse list
H = SparsePauliOp.from_sparse_list([
    ("Z", [0, 1], 0.25),
    ("X", [0, 1], -0.5)
], num_qubits=2)

# Apply layout after transpilation
isa_observable = H.apply_layout(isa_circuit.layout)

5. VQE (Variational Quantum Eigensolver)

from qiskit.circuit.library import TwoLocal
from qiskit_algorithms import VQE
from qiskit_algorithms.optimizers import SLSQP
from qiskit.primitives import Estimator

# Define ansatz
ansatz = TwoLocal(num_qubits=2, rotation_blocks='ry', entanglement_blocks='cx')

# Define Hamiltonian
hamiltonian = SparsePauliOp.from_list([("ZZ", 1.0), ("XX", -0.5)])

# Run VQE
estimator = Estimator()
optimizer = SLSQP(maxiter=100)
vqe = VQE(estimator, ansatz, optimizer)
result = vqe.compute_minimum_eigenvalue(hamiltonian)

print(f"Eigenvalue: {result.eigenvalue}")
print(f"Optimal parameters: {result.optimal_parameters}")

6. QAOA (Quantum Approximate Optimization Algorithm)

from qiskit.circuit.library import qaoa_ansatz
from qiskit_algorithms import QAOA
from qiskit_algorithms.optimizers import COBYLA

# Define cost Hamiltonian
cost_hamiltonian = SparsePauliOp.from_list([("ZIII", 1), ("IZII", 1), ("IIZI", 1)])

# Create QAOA circuit
p = 2  # number of layers
circuit = qaoa_ansatz(cost_hamiltonian, reps=p)

# Run QAOA
qaoa = QAOA(sampler=Sampler(), optimizer=COBYLA(), reps=p)
result = qaoa.compute_minimum_eigenvalue(cost_hamiltonian)

7. Custom Transpilation Pipeline

from qiskit.transpiler import PassManager
from qiskit.transpiler.passes import (
    Optimize1qGatesDecomposition,
    CXCancellation,
    CommutativeCancellation
)

# Build custom pass manager
pm = PassManager([
    Optimize1qGatesDecomposition(),
    CXCancellation(),
    CommutativeCancellation()
])

# Run passes
optimized_circuit = pm.run(qc)

8. Error Mitigation

from qiskit_ibm_runtime import EstimatorV2 as Estimator
from qiskit_ibm_runtime import Options

# Configure error mitigation
options = Options()
options.resilience_level = 1  # 0=none, 1=light, 2=heavy
options.optimization_level = 3

# Execute with mitigation
estimator = Estimator(mode=backend, options=options)
job = estimator.run([(isa_circuit, isa_observable)])
result = job.result()

Qiskit Metal (Quantum Hardware Design)

Overview

Open-source quantum EDA platform for designing superconducting quantum devices. Provides Python API + GUI for automating chip design, integrating with Ansys/GDS tools, and performing electromagnetic analysis.

Installation

git clone https://github.com/qiskit-community/qiskit-metal.git
cd qiskit-metal
pip install -e .

Quick Start

from qiskit_metal import designs, MetalGUI
from qiskit_metal.qlibrary.qubits.transmon_pocket import TransmonPocket
from qiskit_metal.qlibrary.tlines.meandered import RouteMeander

# Create design
design = designs.DesignPlanar()
gui = MetalGUI(design)

# Add transmon qubit
qubit_opts = dict(
    pos_x='0mm', pos_y='0mm',
    connection_pads=dict(
        a=dict(loc_W=+1, loc_H=+1),
        b=dict(loc_W=-1, loc_H=-1)
    )
)
q1 = TransmonPocket(design, 'Q1', options=qubit_opts)

# Add routing
route_opts = dict(
    pin_inputs=dict(start_pin=dict(component='Q1', pin='a'),
                   end_pin=dict(component='Q1', pin='b')),
    total_length='6mm'
)
route = RouteMeander(design, 'route1', options=route_opts)

# Rebuild and visualize
gui.rebuild()
gui.autoscale()

Core Components

ComponentDescription
TransmonPocketPocket-style transmon qubit
TransmonCrossCross-shaped transmon qubit
TransmonInterdigitatedInterdigitated capacitor transmon
RouteMeanderMeandered transmission line
RouteStraightStraight CPW transmission line
ResonatorLumpedLumped element resonator
CapNInterdigitalInterdigital capacitor

Renderers & Analysis

from qiskit_metal.analyses.quantization import EPRanalysis
from qiskit_metal.renderers.renderer_gds import QGDSRenderer

# GDS Export
gds = QGDSRenderer(design)
gds.options['fabricate'] = True
gds.export_to_gds('my_chip.gds')

# EPR Analysis (requires Ansys)
epr = EPRanalysis(design, "hfss")
epr.sim.setup.max_passes = 15
epr.sim.run()
epr.get_frequencies()
epr.get_coupling_matrix()

Ecosystem Packages

PackagePurpose
qiskit-aerHigh-performance local simulators
qiskit-ibm-runtimeIBM Quantum hardware access
qiskit-algorithmsVQE, QAOA, Grover, Shor algorithms
qiskit-optimizationOptimization problem solvers
qiskit-machine-learningQuantum ML classifiers/regressors
qiskit-natureQuantum chemistry/physics
qiskit-experimentsDevice characterization
qiskit-metalQuantum hardware design

Hardware Providers

ProviderPackage
IBM Quantumqiskit-ibm-runtime
IonQ, Rigetti, AWS Braket, Azure Quantum, QuantinuumProvider-specific packages available

Best Practices

  1. Always transpile circuits to ISA format before hardware execution
  2. Use runtime primitives (SamplerV2, EstimatorV2) instead of deprecated execute()
  3. Apply observable layout after transpilation: observable.apply_layout(circuit.layout)
  4. Set optimization_level=3 for hardware to minimize gate count
  5. Use sessions for multiple related jobs to reduce queue time
  6. Enable error mitigation for noisy hardware via options.resilience_level
  7. Vectorize parameter sweeps in primitives for batch efficiency
  8. Prefer SparsePauliOp over dense matrices for Hamiltonians
  9. Use barrier() to prevent unwanted transpiler optimizations
  10. Save runtime credentials once: QiskitRuntimeService.save_account(token=...)

Common Pitfalls

  • Missing ISA transpilation: Runtime primitives require ISA circuits
  • Ignoring circuit layout: Observables must be mapped with apply_layout()
  • Using deprecated execute(): Replaced by primitives in Qiskit 1.0+
  • Not binding parameters: Parametrized circuits need values via PUB or assign_parameters()
  • Mixing simulator/hardware: Check backend.configuration() before execution
  • Exceeding qubit limits: Verify backend.configuration().n_qubits before submission
  • Forgetting measurement: Sampler requires measure_all() or explicit measurements
  • Dense Hamiltonians: Use SparsePauliOp to avoid memory/performance issues

Version Notes

  • Qiskit 1.0+ (Feb 2024): Introduced SamplerV2/EstimatorV2, deprecated execute()
  • Qiskit SDK 2.3 (Jan 2026): Expanded C API, PauliProductMeasurement, Clifford+T transpilation
  • Qiskit Metal 0.5: Breaking changes from pre-0.5; source install required
  • IBM Runtime 0.43+: Required for latest primitive features

References

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.