Research / Prior Art

Prior art comparison matrix

21 systems, libraries, papers, benchmarks, CI templates, and pilots, compared across every dimension relevant to the proposed CI/CD regression testing idea. Cells use YES / NO / PARTIAL / UNKNOWN only; nothing here was assumed.

This is a condensed version (13 of the ~21 columns in the source table) chosen for on screen readability. The complete matrix, including Academic?, Tracing, and Overlap notes for every system, is in prior_art.md.

Reading the matrix: no row is YES across pytest + Cross Version + Cross SDK + Auto Fault Detection simultaneously. QUTest comes closest (YES on CI/CD, Cross Version, Regression Testing) but is explicitly NO on pytest and Cross SDK. pytest-quantum is YES on pytest but only reaches framework level equivalence, and is NO on Cross Version.

Capability coverage

Each system's two axis scores are counted directly from the matrix cells below. CI/orchestration from pytest + GitHub Actions + CI/CD + Cross Version + Cross SDK, testing depth from Regression Testing + Equivalence Checking + Auto Fault Detection + Bug Corpus. Dot size scales with total capability count. No relationship between systems is implied beyond this, the research found none to visualize.

Hover or focus a point for its name and scores

Full matrix

System	Open Source	Regression Testing	Equivalence Checking	Debugging / Tracing	pytest	GitHub Actions	CI/CD	Cross Version	Cross SDK	Bug Corpus	Auto Fault Detection	State Aware / Autonomous	Overlap with proposed work
MQT QCEC Library github.com/munich-quantum-toolkit/qcec SDKs: Qiskit, OpenQASM	YES	NO	YES	NO	NO	PARTIAL	PARTIAL	NO	NO	NO	PARTIAL	NO	Solves the equivalence checking layer only
MQT Debugger Library / tool github.com/munich-quantum-toolkit/debugger SDKs: OpenQASM	YES	NO	NO	YES	NO	PARTIAL	PARTIAL	NO	NO	NO	PARTIAL	NO	Solves single run assertion debugging only
QMon Monitoring tool arxiv.org/html/2512.13422 SDKs: Unknown	UNKNOWN	NO	NO	PARTIAL	NO	UNKNOWN	UNKNOWN	NO	NO	NO	NO	NO	Low, runtime monitoring only
TraceQ Analysis tool arxiv.org/pdf/2508.14533 SDKs: N/A (surface code)	UNKNOWN	NO	NO	YES	NO	UNKNOWN	UNKNOWN	NO	NO	NO	NO	NO	None, hardware / QEC focused
Microsoft QDK Trace Simulator Product Microsoft QDK documentation SDKs: Q# / QDK	PARTIAL	NO	NO	YES	NO	UNKNOWN	UNKNOWN	NO	NO	NO	NO	NO	Low, resource estimation / debugging, single run
CUDA-Q Statistical Assertions Debugging technique arxiv.org/abs/2507.16255 SDKs: CUDA-Q	UNKNOWN	NO	NO	YES	NO	UNKNOWN	UNKNOWN	NO	NO	NO	PARTIAL	NO	Low, single run statistical debugging
Proq Fault localization cited in arXiv:2506.18458 SDKs: Unknown	UNKNOWN	NO	NO	YES	NO	UNKNOWN	UNKNOWN	NO	NO	NO	YES	NO	None, single circuit fault localization
Bloq / AutoBloq Fault localization arXiv:2506.18458 SDKs: Unknown	UNKNOWN	NO	NO	YES	NO	UNKNOWN	UNKNOWN	NO	NO	NO	YES	NO	None, single circuit fault localization (outperforms Proq)
MorphQ Metamorphic testing arXiv:2206.01111 SDKs: Qiskit (tests the SDK itself)	YES	PARTIAL	NO	NO	UNKNOWN	UNKNOWN	UNKNOWN	NO	NO	NO	YES	NO	Partial, tests Qiskit itself, not user projects; not deeply verified
Bugs4Q / Bugs4Q-Robust Bug corpus github.com/Z-928/Bugs4Q SDKs: Qiskit	YES	N/A	N/A	N/A	NO	NO	N/A	N/A	N/A	YES	NO	NO	Reusable corpus, not a CI tool itself
Cross version Bugs4Q replication study Research study arXiv:2606.27124 SDKs: Qiskit (21 versions)	PARTIAL	YES	NO	NO	NO	UNKNOWN	UNKNOWN	YES	NO	PARTIAL	YES	NO	High, closest academic precedent for cross version regression study
QBugs Bug corpus + infra arXiv:2103.16968 SDKs: Unknown (algorithmic)	YES	N/A	N/A	N/A	UNKNOWN	UNKNOWN	N/A	N/A	N/A	YES	NO	NO	Standardization attempt for debugging technique comparison
QBugLM Agentic benchmark arXiv:2606.07314 SDKs: OpenQASM 3 (framework agnostic)	UNKNOWN	NO	NO	NO	NO	UNKNOWN	UNKNOWN	NO	NO	YES	YES	YES	Closest agentic system found; no CI/CD, no cross version
Qolumbina Benchmark suite arXiv:2607.02029 SDKs: Multiple (curated programs)	YES	NO	NO	NO	UNKNOWN	UNKNOWN	NO	NO	NO	PARTIAL	NO	NO	Benchmark corpus only, not a testing tool
QUTest Testing framework arXiv:2605.19736 SDKs: Qiskit only	YES	YES	NO	NO	NO	YES	YES	YES	NO	NO	PARTIAL	NO	HIGH, the most directly overlapping system found
Benchpress Benchmark suite github.com/Qiskit/benchpress SDKs: Qiskit, Braket, Cirq, Tket, BQSKit, Staq, pyqpanda3, Qiskit IBM transpiler	YES	NO	NO	NO	YES	NO	NO	NO	YES	NO	NO	NO	Proves the pytest native cross SDK harness pattern at scale, but measures capability/performance only; detects no regressions or equivalence failures
pytest-quantum pytest plugin github.com/qbench/pytest-quantum SDKs: Qiskit, Cirq, pytket	YES	PARTIAL	PARTIAL	NO	YES	YES	YES	NO	YES	NO	PARTIAL	NO	Partial, pytest native + cross framework equivalence, but no cross version
qtest-quantum pytest plugin github.com/metin-5115/qtest SDKs: Qiskit (implied)	YES	PARTIAL	PARTIAL	NO	YES	YES	YES	NO	NO	NO	PARTIAL	NO	Partial, pytest native, single SDK only, immature (0 stars)
qc-assert pytest plugin github.com/JMORAF87/qc-assert SDKs: Unknown	YES	PARTIAL	NO	NO	YES	YES	NO	NO	NO	NO	NO	NO	Minimal, intent stated, not implemented
Qiskit ecosystem CI template CI template github.com/Qiskit/qiskit-addon-cutting SDKs: Qiskit	YES	PARTIAL	NO	NO	UNKNOWN	YES	YES	YES	NO	NO	NO	NO	Reusable version tier CI, but no behavioral diffing or regression detection
Qiskit QPY compat harness Internal CI github.com/Qiskit/qiskit (.github/workflows/qpy.yml) SDKs: Qiskit	YES	YES	NO	NO	UNKNOWN	YES	YES	YES	NO	NO	YES	NO	Narrow, serialization compatibility only, not externally reusable
quantum-transpiler-regression-testing (“cart”) Research tool / pilot github.com/furqan-nr/quantum-transpiler-regression-testing SDKs: Qiskit transpiler only	YES	YES	PARTIAL	NO	NO	UNKNOWN	UNKNOWN	NO	NO	NO	YES	NO	High, closest regression testing pilot; single SDK, CLI, brand new, 0 stars

Full 21-column matrix and per system narrative: research/prior_art.md in the GitHub repository