Known limitations

Lab makes planning, readiness, custody, observations, and promotion reviewable. It does not control physical instruments, guarantee upstream recommendation quality, or turn a well-formed record into a complete scientific answer.

Operational and scientific limits

Limitation	Consequence	Responsible interpretation
physical bench and instrument execution occur outside the package	handoff success is not proof that work ran as instructed	require returned observation, custody, and deviation evidence
readiness evaluates declared resources and evidence	undeclared staffing, maintenance, contamination, logistics, or local policy can still block work	record the site-specific readiness inputs and review time
LIMS and vendor targets differ in representational power	export can flatten or omit constraints even when serialization succeeds	publish mapping, validation, and loss reports
assay design uses explicit assumptions and available context	power advice and controls can be wrong for unmodeled variance or biology	preserve assumptions and require domain review
queue and scheduling models cover declared capacity and compatibility	real disruptions can invalidate an otherwise stable schedule	re-evaluate readiness before handoff
returned observations depend on external measurement and custody quality	a precise stored value can still be biased, mislabeled, or incomplete	inspect controls, lineage, deviations, missingness, and QC
acceptance rules are assay-specific	passing technical QC does not establish biological causality or utility	separate technical acceptance from downstream interpretation
promotion policy operates on available evidence	later contradiction can weaken a promoted claim without erasing history	append the new evidence and reconciliation record

Boundary of a successful record

flowchart LR
    H["valid handoff"] -. not proof of .-> X["physical execution"]
    X --> O["returned observation"]
    O -. not proof of .-> A["assay acceptance"]
    A -. not proof of .-> B["biological conclusion"]
    B -. not proof of .-> U["practical utility"]

Each transition needs distinct evidence. The package preserves those transitions so callers do not infer the next state from the previous one.

Report the operating envelope

Name site and time context where relevant, readiness inputs, controls, authority, external system and mapping version, observation completeness, deviations, QC rule, reliability, and promotion policy. When physical execution evidence is absent, describe the artifact as a plan or handoff—not a completed experiment.