Flagship Release Candidate¶

This page names the strongest current release sentence the repository can defend. It does not describe the whole repository as universally ready.

The current flagship release candidate is:

outsider-auditable, bounded: dda, dia, ptm, targeted
review-grade, bounded: lfq
internal support only: multiplex

That bundle is materially stronger than it was at v0.3.7, but it is still a bounded family bundle, not a blanket claim about all proteomics workflows.

What Makes This A Real Bundle¶

The release-candidate sentence now depends on a deeper cross-package chain:

benchmark packages that expose real scientific evidence roots
runtime rerun lanes and artifact-boundary checks
grounded claim and contradiction review
recommendation challenge, downgrade, and regret surfaces
downstream lab-consequence and refusal pressure

If one of those layers narrows, the bundle narrows with it.

Current Bundle Inventory¶

family	current public release sentence	why it survives	why it still stays bounded
`dda`	outsider-auditable, bounded	strong benchmark packet, downstream review surfaces, and cross-package consequence route	still `import_only`, so runtime realism remains narrower than full live-engine parity
`dia`	outsider-auditable, bounded	raw-executable runtime lane plus stronger DIA-specific scientific and consequence surfaces	library incompleteness and consequence ambiguity still block broader language
`lfq`	review-grade, bounded	public benchmark, runtime, grounding, and recommendation surfaces are real and reviewable	missingness, normalization, cohort transfer, and external review pressure still narrow the stronger sentence
`ptm`	outsider-auditable, bounded	stronger localization, review, and runtime surfaces	localization strength still outruns downstream consequence confidence
`targeted`	outsider-auditable, bounded	strong benchmark and runtime packet plus explicit challenge and lab surfaces	calibration, interference, and follow-up burden still block stronger authority
`multiplex`	internal support only	real package and runtime substance exists	the current stress packet still collapses outsider-facing trust

What This Bundle Proves¶

the repository can now defend multiple outsider-auditable workflow families without pretending that they are all equally strong
the strongest release language now rests on scientific, runtime, grounding, recommendation, and consequence surfaces together rather than on release prose alone
the release sentence is now specific enough that a skeptical reviewer can challenge each family through shipped public routes

What This Bundle Still Refuses To Claim¶

repository-wide excellence or universal proteomics authority
decision-grade trust for every public family
full family equivalence just because several families now survive bounded outsider audit
stronger follow-up confidence than the current lab-consequence chain supports

Best Audit Route¶

Use this order when challenging the current flagship sentence:

Why The Bundle Still Stays Narrow¶

strong family packets do not erase cross-family transfer limits
one raw-executable lane does not erase downstream assay-burden limits
broader scientific depth in core does not erase narrower runtime, grounding, or consequence boundaries
if the weakest family-limiting surface still survives public scrutiny, the release sentence must stop there

Honest Reading¶

The practical release meaning is simple:

dda, dia, ptm, and targeted are public and challengeable now, but each still carries a visible limiting factor
lfq is publicly reviewable and materially stronger than before, but its present release wording still has to remain narrower than the outsider call
multiplex is real product substance, not fake scaffolding, but the repository still cannot let it inherit trust from neighboring stronger families

Boundary¶

This page names the strongest current release sentence. It does not authorize broader wording than the family, runtime, grounding, and consequence surfaces can all still defend together.