Developer builds three Quantinuum Guppy tools, finds bugs

Developing tools for the emerging quantum computing ecosystem requires rigorous attention to detail. Recently, a software developer built utility tools for guppylang, Quantinuum's Python-embedded quantum programming language that compiles to HUGR and runs on their Selene simulator and trapped-ion hardware.

The primary focus was ensuring absolute correctness rather than just building simple demos. Every formula was cited to its source, and compiler behavior was verified through actual code compilation and output inspection rather than relying blindly on documentation.

Stock photo for illustration only, not from the actual event

This strict discipline led to the creation of three distinct developer tools:

• qshelf: A tested package registry of quantum algorithm implementations (QFT, Grover, QAOA, VQE-H2) verified against independent mathematical references.
• Estimand: A fault-tolerant resource estimator adapter that extracts gate-count summaries from real guppy control flows and feeds them to Qualtran's cost models.
• qmatchpoint: A utility that connects PyMatching to syndrome bits produced by a guppy QEC circuit, filling a current gap in the stack.

Finding bugs was never the initial goal, but insisting on strict verification uncovered four issues within guppylang itself. These included a wrong unitary from standalone versus combined iqft and qft compilation, a flawed unitary from multi-controlled Z, a rejected generic array-length type, and a rejected numpy.ndarray closure.