Integrate a new decoder into LightStim's decoder backend so it works with `SimulationPipeline` and `DecoderConfig`. Use this skill whenever the user asks to add a custom decoder, wrap a third-party decoder library (e.g. a neural decoder, a research-paper decoder), register a CPU or GPU decoder backend, build a decoder from scratch using a DEM, or extend the existing BPOSD/PyMatching/MWPF stack with a new variant.
Run a LightStim circuit through the simulation pipeline and extract logical error rates. Use this skill whenever the user asks to simulate a circuit, get LER vs physical error rate, run PyMatching or BPOSD or MWPF decoding, sweep noise parameters, benchmark a code's threshold, plot logical error rates, configure post-selection (for state injection or distillation), or target specific logical observables in a multi-qubit circuit.
LightStim known pitfalls, bugs, and non-obvious API contracts. Read this whenever a build produces unexpected results, the tracker raises a RuntimeError, noiseless-check detections appear, LER is suspiciously high, or the user reports "it was working and now it isn't." Each entry states the symptom, the root cause, and the fix pattern.
Direct use of CircuitBuilder and SyndromeTracker to construct a custom QEC circuit from scratch — without a pre-built experiment class. Use this skill whenever the user wants to: implement a novel protocol from a paper, chain multiple logical operations with full control over each step, build a circuit that no existing XXXExperiment class covers, or understand how detectors and observables are generated internally.
Configure noise models and error rates for LightStim experiments. Use this skill whenever the user asks about noise models, setting physical error rates, choosing between circuit_level / phenomenological / code_capacity / XZ_biased noise, biased noise for neutral atom or trapped ion hardware, or wants to understand how noise is applied to a circuit. Also trigger when the user asks "what p should I use?" or "how do I add noise to my circuit?"
Add a new QEC code to LightStim by implementing QECPatch and a syndrome extraction block. Use this skill whenever the user asks to implement a new quantum error correcting code, extend LightStim with a custom code family, define stabilizers and logical operators for a new code, create a new SE_block, or understand the minimal interface a code needs to satisfy.
Design a new LogicalCouplerProtocol in LightStim to implement inter-patch joint measurements (lattice surgery). Use this skill when the user wants to: implement a new logical coupler for a code that doesn't have one yet, understand how the Unrotated Surface Code coupler works internally, design a coupler for a new code family (e.g. BB code logical measurement), or extend an existing coupler with a new boundary configuration. This skill is the key to replicating papers that describe custom "logical coupler" constructions without open-sourcing their implementation.