Contract field reference¶
Choose a supported contract below. Read the reference guide for how inputs, run indexes and engineering results fit together.
Project inputs¶
- Mission inputs: Describe the operational environment, modes, lifetime and engineering statements that the design must address.
- Spacecraft inputs: Describe payloads, chosen subsystem families, catalogs and physical accommodation policies.
- Study inputs: Connect one run to mission and spacecraft inputs. Multi-case experiments use a separately confirmed Design Studies plan.
- Resource references: Name a runtime resource or a project-local resource with optional content-integrity verification.
- Engineering statements: Record the criterion, applicability and evidence behind a requirement, acceptance condition or assumption.
- Evidence sources: Record provenance and confidence for authored inputs and engineering statements; source attribution does not automatically establish validation.
Run and result contracts¶
- Run manifest: The authoritative entry point to a run bundle. Follow its indexed references and verify the recorded lifecycle.
- Result index: Discover registered results through the manifest-selected index and inspect the actual available records.
- Artifact index: Discover registered files by artifact identity, role and recorded integrity data.
- Result envelope: The common wrapper for a domain result. Interpret execution, engineering assessment, applicability and issues before using payload values.
- Issues and diagnostics: Inspect issue code, severity, message, affected statements and action hints before choosing a remedy.
- Candidate ledger: Review evaluated candidates and their recorded outcomes without inferring global optimality from the selected candidate.
- Selected design: Inspect the design materialized by the study and its links to the candidate and result evidence.
- Convergence iteration: Read the iteration state and coupling observations; a terminal execution status alone does not establish convergence.
Engineering results¶
- Spacecraft mass properties: Inspect mass, center of gravity and inertia in the declared frames and configurations; do not mix stowed and deployed states.
- Mass results: Compare the supported mass cases and selections. Distinguish nominal convergence dry mass, budget margins and wet mass.
- Power and EPS results: Inspect generation, load, storage, degradation, margins and applicability together.
- Lifetime results: Review supported lifetime outcomes and the assumptions linking operating demand, storage and degradation.
- Thermal results: Review thermal state, heat-flow and heater/radiator evidence together with assumptions and convergence.
- Propulsion results: Distinguish authored maneuver allocations, sizing assumptions and derived propulsion/propellant results.
- Communications results: Review link and contact assumptions, selected hardware, availability and data-service evidence separately.
- ADCS results: Interpret sizing and pointing results within the declared disturbance, hardware and operational assumptions.
- ADCS robustness results: Compare the recorded robustness cases and their limits; a sweep is not independent flight validation.
- Launch accommodation results: Review the explicit launch profile and its geometric, wet-mass, center-of-gravity, loads and frequency screening outcomes.
- Environment results: Inspect the environment and operating quantities used by downstream calculations within their supported assumptions.
- Geometry loading results: Inspect the authoritative geometry/environment loading passed to dependent analyses.
- Placement results: Review component placement, validity and constraints within the selected geometry/topology.
- Deployable assembly: Inspect resolved assembly/configuration data when that artifact is actually present; fixed-panel cases need not emit it.
- Cost projection: Inspect cost-policy inputs and estimate authority. Cost output is carried by the result envelope; no standalone cost-result schema is published.
Evaluation and advisory review¶
- Evaluation manifest: Verify source identities, policy snapshot and immutable evaluation outputs.
- Evaluation policy: Inspect approved preference mappings, category weights, caps, applicability and materiality.
- Authorized policy overrides: Record sparse policy changes with captured human authorization; agents cannot authorize changes.
- Candidate evaluation: Read mandatory outcomes, score completeness, KPI evidence, categories and limiting caps separately.
- Candidate comparison: Inspect compatibility, Pareto dominance and proposed co-leaders without changing selection or approval.
- Evaluation presentation: Consume a versioned scorecard projection with source-linked detailed evidence.
- Neutral review evidence: Read bounded captured evidence and provenance without P05 scores or recommendations.
- Review revision manifest: Verify append-only review state, source bindings and predecessor hashes.
- Blind advisory review: Preserve independent evidence-cited findings before deterministic assessment disclosure.
- Advisory reconciliation: Retain agreements, disagreements, corrections and unresolved issues beside the frozen findings.