Qloop

A quantum-circuit SDLC framework built on classical simulators: a plugin registry that auto-discovers circuits, six pipeline stages (lint → exact → property → transpile → noisy → hardware), and a 17-circuit corpus that exercises every stage.

If you’ve never seen this repo before, read these in order:

  1. Architecture — how the pieces fit together: the CircuitSpec contract, the registry, the six stages, and where Stim fits in.
  2. Installation — get a working environment.
  3. Usage — run the pipeline, run one stage, read the metrics artifact.
  4. Testing — how the test suite is organized, and how to debug one circuit.
  5. Adding a circuit — the only page you need to contribute a new circuit.
  6. Extending the framework — changes that go deeper than “add a circuit”: new invariants, new backends, new contract methods.
  7. Circuit corpus — reference table of all 17 registered circuits.

The one-sentence pitch

Drop a single Python file implementing a small contract (CircuitSpec) into qloop/circuits/, and it is automatically picked up by every test tier — exact verification, property-based invariants, a transpilation budget check across multiple hardware topologies, and (optionally) noisy-simulation tolerance bands — with zero edits to the test suite, CI config, or any registration list.

# qloop/circuits/my_circuit.py
from qiskit import QuantumCircuit
from qloop.core.spec import CircuitSpec, Budget

class MyCircuitSpec(CircuitSpec):
    name = "my-circuit"
    n_qubits = 2

    def build(self, **params) -> QuantumCircuit:
        qc = QuantumCircuit(2)
        qc.h(0)
        qc.cx(0, 1)
        return qc

    def budget(self) -> Budget:
        return Budget(depth=10, two_qubit_gates=5)

pytest tests/generic/ -v -k my-circuit

That’s it — no import list, no CI edit, no test file to write. See Adding a circuit for the full contract, including which optional methods unlock which pipeline stages.

Source

github.com/sangmeshcp/Qloop