contract_id	label	product_role	primary_question	required_sections	required_links	current_route	current_state	next_gate	inspired_by	claim_boundary
PRODUCT-CONTRACT-DATASET-CARD-1.0	Dataset Card	Turn a stored matrix or source accession into a citable, task-qualified biological data object.	What was measured, in which biological units, and what can this dataset validly support?	Identity and version | Disease, tissue, cell state and perturbation context | Sample, donor and statistical-unit design | Matrix, feature, metadata and checksum inventory | Access, licence and source provenance | Readiness, permitted use and prohibited use | Linked ModelRuns, benchmarks and studies	source record | download or access object | readiness gate | downstream run or study	/resource/datasets/	RELEASED_PARTIAL_STANDARDIZATION	Render the same required sections for every dataset and reject model-ready status when donor, feature or perturbation identity is unresolved.	CELLxGENE | Tahoe-100M | DepMap	Dataset scale or availability does not establish independent biological replication or task suitability.
PRODUCT-CONTRACT-MODEL-CARD-1.0	Model Card	Bind a model identity to one task, one execution history and one bounded claim surface.	What did this exact model version receive, return, beat or fail, and where must it abstain?	Model identity, version and provider | Task, input and output object | Training and evaluation context | Split, leakage and baseline controls | Metrics, seeds and local execution result | Failure, abstention and prohibited claims | Artifacts, run receipts and reproducibility links	ModelRun | dataset digest | benchmark task | artifact or source | failure analysis	/resource/models/	RELEASED_AUDITED	Expose a uniform adapter and quickstart only after the same model identity is reproducible across documentation, API and run receipts.	CZI Virtual Cells Platform | X-Cell | NVIDIA BioNeMo	Architecture reputation and external performance are not local NMD or DMD validation.
PRODUCT-CONTRACT-BENCHMARK-CARD-1.0	Benchmark Card	Make model evaluation a frozen scientific task rather than a movable leaderboard claim.	On which sealed outcomes, split, baselines and metrics did a model advance or stop?	Scientific question and task identifier | Frozen data, target context and outcome object | Split, leakage groups and hidden-label policy | Permanent simple baselines | Metric families, seeds and uncertainty | Advance or no-advance decision gate | Immutable result and negative-run history	benchmark schema | split manifest | baseline runs | ModelRuns | result download	/resource/benchmarks	RELEASED_TASK_BOUNDED	Add external adapters only on compatible frozen tasks and retain every stopped or negative result.	Arc Virtual Cell Challenge | CZI Virtual Cells benchmarks	A same-assay benchmark result cannot be transferred to DMD muscle or a different output distribution.
PRODUCT-CONTRACT-WORKFLOW-OUTCOME-CARD-1.0	Workflow / Outcome Card	Connect a biological question to a frozen experiment, returned outcome and explicit evidence update.	What experiment can change the conclusion, and how will every positive, null, toxic or failed outcome return?	Question, estimand and hypothesis | Intervention, cell context, dose, time and comparator | Biological unit, controls, endpoints and QC | Lifecycle, preregistration and immutable decision rule | Prediction freeze when applicable | Measured outcome, provenance and failure state | Evidence transition and claim update	Study Card | Prediction Card when applicable | Outcome object | source evidence | release update	/resource/planner	SCHEMA_RELEASED_OUTCOME_REGISTRY_EMPTY	Register one immutable DMD study and return its outcome through the same identifiers without suppressing null, toxic or failed branches.	Recursion OS | Cellular Intelligence | VCell	A study design is prospective infrastructure, not a measured intervention effect.