The interface is localized. Skill descriptions and documents are machine-translated by AI into your language; code, commands, paths and technical terms stay in the original, and the author's document governs.
Explore skills
Development · SKILL.md

core-circuits-and-ops

core-circuits-and-ops

Build, inspect, mutate, serialize, and debug Cirq circuits and core operation objects.

Independent GitHub snapshot · Automated classification · Original documentation governs

Skill documentation

GitHub

name: core-circuits-and-ops description: "Build, inspect, mutate, serialize, and debug Cirq circuits and core operation objects." disable-model-invocation: true metadata: disco-role: operating license: Apache 2.0

Core Circuits and Operations

Use this sub-skill when the task is about Cirq's core object model: qubits/qids, gates, operations, moments, circuits, circuit diagrams, measurement keys, tags, controls, symbolic parameters, custom gates, and basic object serialization.

Stay in this sub-skill for

  • Choosing or constructing LineQubit, GridQubit, NamedQubit, LineQid, or GridQid objects.
  • Building and mutating Gate, Operation, Moment, Circuit, FrozenCircuit, and CircuitOperation objects.
  • Using Circuit.append, Circuit.insert, InsertStrategy, OP trees, findall_operations, measurement keys, operation tags, controls, and classically controlled operations.
  • Resolving symbolic parameters with ParamResolver and resolve_parameters.
  • Using core protocols such as unitary, has_unitary, decompose, inverse, qid_shape, and circuit_diagram_info for debugging custom or built-in objects.
  • Basic JSON, QASM, and Quirk interop at the circuit-object level.

Route away when

  • The user needs sampling, state vectors, density matrices, result histograms, parameter sweeps as execution, or noise simulation: use simulation-study-and-noise.
  • The user needs circuit transformers, optimization, decomposition pipelines, routing, target gatesets, or placement: use transformers-and-compilation.
  • The user needs cloud/hardware provider serializers, live services, credentials, or provider-specific package behavior: use hardware-providers-and-serialization.
  • The user needs algorithm recipes, Pauli observables, expectation values, or textbook algorithm interpretation: use algorithms-and-observables.

Operating routine

  1. Identify whether the task is object construction, mutation, inspection, parameters, protocols, or interop.
  2. For exact signatures and object-model facts, read references/api-reference.md.
  3. For common build/debug recipes, read references/workflows.md.
  4. For known failure modes, read references/troubleshooting.md.
  5. For a safe local smoke helper, run scripts/inspect_circuit.py with --help, then run it with optional --json or --qubits N.

Bundled assets

  • references/api-reference.md — verified public API signatures and compact examples for core Cirq objects.
  • references/workflows.md — task-oriented recipes for construction, mutation, parameter resolution, custom gates, and interop.
  • references/troubleshooting.md — symptom-based fixes for unresolved parameters, duplicate measurement keys, protocol failures, qid-shape mismatches, JSON resolver issues, QASM limits, and insert strategy surprises.
  • scripts/inspect_circuit.py — deterministic helper that builds a small parameterized circuit, prints diagrams, validates measurement-key uniqueness, and optionally performs a JSON roundtrip.
Saved skills