Workflow Families

This page is the fastest route from a workflow-family label to the sentence the repository can honestly publish today.

The family table is not a marketing summary. It is the place where scientific breadth, runtime realism, grounding pressure, and downstream consequence are collapsed into one bounded family call.

How To Read The Table

• trust status tells you the strongest sentence the repository can publish today for that family
• primary run mode tells you whether the strongest current route is raw_executable or still import_only
• scientific center of gravity tells you what kind of scientific burden is most visible in the family
• benchmark coverage tells you where a serious reviewer should start
• current blockers tells you why stronger language is still blocked

Family Comparison

family	trust status	primary run mode	scientific center of gravity	benchmark coverage	current blockers
dda	outsider-auditable, bounded	import_only	reviewed downstream replay and identification-facing evidence	DDA Benchmark Lineage	in-repo live-engine parity still remains weaker than the reviewed downstream path
dia	outsider-auditable, bounded	raw_executable	library-aware DIA evidence, replay, and reviewable runtime proof	DIA Benchmark Lineage	library incompleteness and downstream absent-peptide consequence still narrow the broader sentence
lfq	review-grade, bounded	raw_executable	normalization, missingness, and cohort-sensitive quantification review	LFQ Benchmark Lineage	missingness, normalization pressure, and external-review-kit limits still block outsider-auditable release language
multiplex	internal support only	raw_executable	runtime substance exists, but public trust still fails under stress	Multiplex Benchmark Lineage	the current stress package still collapses outsider-facing trust
ptm	outsider-auditable, bounded	raw_executable	PTM localization, site review, and consequence-sensitive follow-up	PTM Benchmark Lineage	localization evidence is stronger than downstream consequence confidence
targeted	outsider-auditable, bounded	raw_executable	targeted panels, calibration, interference review, and validation planning	Targeted Benchmark Lineage	calibration, interference, and assay-burden pressure still block broader certainty

Why The Families Are Substantively Stronger Now

The family labels now sit on top of deeper package substance than they did at v0.3.7:

• core carries broader sequence, chemistry, spectra, mzML, DIA, PTM, and quantification surfaces
• runtime carries replay, rerun, refusal, and artifact-integrity routes
• knowledge carries workflow-level claim grounding and contradiction review
• intelligence carries recommendation challenge, downgrade, and regret surfaces
• lab carries control demand, refusal, and requested-versus-observed outcome loops

That deeper product is exactly why the family calls have to stay explicit. More real substance makes overclaiming easier if the limiter is not named.

What The Family Calls Refuse To Hide

• a strong benchmark package does not erase runtime weakness
• a reproducible rerun lane does not erase grounding or consequence weakness
• a strong recommendation packet does not erase assay-burden pressure
• one strong family packet does not authorize broader repository-wide language

How To Challenge A Family Sentence

• start at the lineage page to inspect what evidence roots are actually public
• open runtime execution pages to confirm the current run mode is real
• open grounding and recommendation pages to see whether the sentence narrows after contradiction or challenge
• end at lab consequence when the real limiter is downstream burden rather than analytical elegance

Best Next Questions

• Open DDA Cross-Package Handbook when you want one worked example of how a family route crosses benchmark, runtime, grounding, recommendation, and consequence.
• Open Flagship Release Candidate when the question is how the current family calls combine into one release bundle.
• Open Scientist Journey when the question is how one careful reader should inspect a family from benchmark packet to downstream consequence.
• Open Benchmark Assets when the question is whether the public evidence root is broad enough and honest enough.
• Open Execution when the question is whether the runtime lane still deserves the published sentence.
• Open Decision Support when the question is whether grounding, recommendation posture, or lab burden is the real limiter.

Reader Rule

If the family table sounds stronger than the family-specific trust page, rerun surface, or consequence route, the table is wrong and must narrow.

Boundary

This page compares family-level public posture. It should hand the reader to the evidence, runtime, grounding, recommendation, or lab owner once the real question is known.