Make the next
observation count.

Next required test. Bind every observation to the implemented operator; separate fixed-point error and sampling error; retain the recurrence model-class assumption.

Next required test. Bind a real reference-channel calibration and intervention controls; preserve non-identifiability and record covariance shared across channels.

Next required test. Preserve the exact generator assumptions and the common 152-formula check. Expand to reserved harder families with matched timing/resource budgets before claiming a branch improvement.

Next required test. Add spectral line-spread, exposure and polarization response; reserve independently sourced transition/intensity targets with covariance.

Next required test. Add wavelength-response/illuminant calibration, material temperature and uncertainty; predict reserved reflectance or diffraction frames rather than just an RGB rendering.

Next required test. Current 0.25.0 reconstruction is runnable and tested. Preserve original-byte losses; independently bind physical units/coupling, calibrate apparatus and reserve new measurements.

Next required test. Keep inspected frames as development data; specify an open-propagation microscope baseline and reserve fresh frames before fitting it.

Next required test. Preserve the holdout with no claimed improvement; audit shared theory/experimental covariance and reserve a new transition before changing the operator.

Next required test. Keep setting chronology, optional-stopping guarantees and composite Born-null calibration; test new hardware records only through the hardware protocol.

Next required test. Use this evidence classification as a common policy; do not promote an unsupported residual or a simulator pass to a physical wall break.

Next required test. Measure the reference/sample path mismatch and missed-event fractions independently; predict a newly reserved time-arrival record with those uncertainties.

Next required test. Obtain independent in-sample optical and recording-response calibration and genuinely new trial-level biological records before fitting or claiming a unique mechanism.

Next required test. Independently justify simultaneous distance bounds, synchronization and metric premises; challenge curved/routed 3D alternatives on new records.

Next required test. Characterize pulse/threshold or spectral-resolution law independently; separate resolution variable from physical time and test an intervention that breaks the alias.

Next required test. Use its fail-closed provenance pattern elsewhere; actual QPU data require a supported backend, independent calibration envelope and untouched trials; do not synthesize hardware evidence.

Next required test. Keep the exact rule universe and independent certificate; any speed claim needs matched hardware, repeated timing and held-out positions.

Next required test. Bind rule/history semantics and exact certificates; separate search counts from runtime uncertainty.

Next required test. Recheck every pruning claim with an independent exhaustive oracle on reserved small positions.

Next required test. Never replace legal minimax conditions with a visual wave score; use fresh oracle positions and account for proof cost.

Next required test. Keep ordering separate from pruning; test frozen ordering against matched baseline on new positions.

Next required test. Freeze corpus, rule history and timing environment; score independent certificates and end-to-end resource cost.

Next required test. Report horizon and unresolved intervals; reserve larger/deeper fixtures and repeat controlled timings.

Next required test. Aggregate verifier 1.1 and all 17 case replays plus six repeats pass. Preserve missing old individual-file history; assess reserved positions and controlled training-inclusive performance before a new improvement claim.

Next required test. Acquire an independently characterized profile/tail bound; keep calibration certificate conditional and test reserved temperatures/response profiles.

Next required test. Independently map the energy readout to a physical instrument and restrict ordinary-space competitors; uncertainty must include calibration shared across vertices.

Next required test. Fit Hamiltonian/readout efficiencies only to independent calibration; reserve actual probe outcomes and test competing instrument/state models.

Next required test. Add independently calibrated pulse, dephasing, readout and sampling errors; predict reserved time-correlation distributions rather than ideal expectation curves.

Next required test. Use independent light/geometry calibration and sensor response; reserve new material/lighting conditions; no optical fit certifies a theological reading.

Next required test. Freeze textual coding before new passages; physical tracer proposals need apparatus/readout/noise models and physical validation, not verse selection.

Next required test. Add flux, cross-sections, efficiency and energy-migration response with independently bounded nuisance parameters and reserved event data.

Next required test. Verify metric, curvature and calibration externally; reserve metric-distance records and retain routed or mirror alternatives.

Next required test. Calibrate clock drift, frequency dispersion and path stability independently; validate on a reserved physical geometry with shared error bounds.

Next required test. Model full pulse waveform/detector tails and obtain a controlled source/reference path experiment; a centroid-only fit cannot identify the cause.

Next required test. Independently calibrate dark counts, thermal leakage, readout work and efficiency; predict a physical blank/control before calling an event unexplained.

Next required test. Measure preparation/readout drift and calorimetry independently; keep decision-rule selection frozen and share calibration covariance across waits.

Next required test. Validate assignment isolation, timing and physical readout calibration; preserve fixed-horizon scope and avoid post-hoc rule selection.

Next required test. Retire proxy matches as evidence of a solution; use a full common background/likelihood and predict a quantity not used to select the amplitude.

Next required test. Preserve measured covariance and calibrated-vs-tested target split; independently evaluate sensor/summary likelihood assumptions and test newly reserved summaries.

Next required test. The eight-case source rebuild and exact analysis comparison are complete. Continue with the official correlated likelihood and independent probes; do not use the archived display-point surrogate as that likelihood.

Next required test. Full native likelihood and four refined solver replays pass. Preserve nonstationary/nonunique endpoints; investigate optimizer/posterior convergence and reserve independent probes before changed-model validation.

Next required test. Source lineage and the small common corpus are reconciled. Test larger reserved families under the same oracle and resource rules; v27 and v28 share the same counting core.

Next required test. Preserve the inversion bijection and exact common-corpus matches. Assess total representation/search cost on larger reserved families; the finite result proves neither a universal speedup nor P=NP.

Next required test. Keep finite certificates, complexity bounds, hard negative families and common benchmark resource accounting.

Next required test. Evaluate neural weights separately from engineered solvers; use equal observation/tick budgets, untouched generators, seed variation and training-inclusive compute.

Next required test. Keep a linear/global control and equal parameter/compute/data budgets; add reserved graph sizes, multiple seeds and prediction calibration before any frontier-scale claim.

Next required test. Add finite source/detector transfer and material/geometry uncertainty; predict independently measured scattering while keeping discretization error separate from absorption.

Next required test. Add calibrated spatial/temporal response before physical timing claims; require invariance of inferred physics under changed observation resolution.

Next required test. Reserve material/propagation data and independently characterize the constitutive law before treating a geometric mesh as a physical absorber.

Next required test. Propagate source uncertainty and do not extrapolate to cosmic temperatures; reserve spectra at a specified material/temperature/geometry.

Next required test. Measure response matrix, pose and phase calibration, bound unmodelled modes, then predict reserved motion/vector observations.

Next required test. For each retained route, identify a calibrated physical observable and an intervention excluding its specific surviving countermodel.

Next required test. Characterize real camera pose, lens response, occlusion and timing; validate prediction-only loss and recovery on reserved multi-camera drop tracks.

Next required test. Independently calibrate light directions/intensity, reflectance and camera response; score unseen geometry, not just image agreement.

Next required test. Retain exact object/projection semantics; no physical measurement claim applies without a new physical model.

Next required test. Freeze linguistic coding and alternatives; any extracted physical conjecture needs its own measured protocol and cannot inherit empirical status from the text.

Next required test. Expose units/approximation scope; a rendered visual match must not substitute for a calibrated radiance/camera prediction.

Next required test. Archive as an unsupported historical conversation proposal. Do not import its flip-rate, power, radius or quartz numbers into the current manuscript evidence.