Scope and non-goals

bijux-proteomics-lab owns the contracts that turn an evidence need into reviewed laboratory work and return the observed consequence. It does not operate physical instruments or own the scientific and decision policies that precede and follow that work.

Owned lifecycle

flowchart LR
    E["evidence need"] --> P["assay and experiment design"]
    P --> R["readiness, batching, and scheduling"]
    R --> H["authorized handoff and custody"]
    H --> X["external physical execution"]
    X --> O["observation, QC, and reliability"]
    O --> C["reconciliation, promotion, and follow-up"]

The package owns the solid-arrow records around external execution: design, dependencies, readiness, queueing, scheduling, authority, custody, export, observation intake, QC, failure classification, reconciliation, and promotion posture.

Explicit non-goals

Not owned here	Responsible boundary
physical instrument control, acquisition software, or bench automation	laboratory systems and operators outside this Python package
generic repository workflow execution and run-state infrastructure	Bijux Proteomics Runtime
scientific transformations, scoring, quantification, and workflow-family acceptance	Bijux Proteomics Core
source, citation, claim, and contradiction custody	Bijux Proteomics Knowledge
ranking, recommendation, sensitivity, regret, and advisory policy	Bijux Proteomics Intelligence
shared identifiers, outcomes, canonical serialization, and migration primitives	Bijux Proteomics Foundation
repository policy, CI, release, and documentation automation	Bijux Proteomics Maintain

Boundary rules

• an advisory plan remains non-executable until readiness and authority are recorded for a concrete batch;
• an authorized handoff records intended work but does not claim that physical execution occurred;
• an observation records what returned and does not imply QC acceptance;
• an accepted assay does not automatically become promoted evidence;
• promotion appends a downstream disposition and never rewrites the plan or observation;
• failures, refusals, deviations, and inconclusive results remain first-class outcomes.

Use the execution model for the state machine and the operator workflow for the end-to-end handoff and return path.