@researai/deepscientist 1.5.16 → 1.6.0

package/src/skills/science/references/packages/maxwelllink.md

@@ -0,0 +1,80 @@
+# MaxwellLink Light-Matter Co-Simulation
+
+## Catalog
+
+- Package id: `maxwelllink`
+- Domains: `molecular_dynamics`, `electromagnetics`, `workflow_provenance`
+- Tags: `light-matter`, `electrodynamics`, `multiphysics`, `molecular-dynamics`, `quantum-dynamics`, `hpc`
+- Knowledge URL: https://github.com/skilled-scipkg/MaxwellLink
+- Source archive URL: https://github.com/skilled-scipkg/MaxwellLink/archive/refs/heads/main.zip
+- Upstream project URL: https://github.com/TaoELi/MaxwellLink
+- Homepage: https://taoeli.github.io/MaxwellLink/
+- Catalog source: FermiLink skilled-scipkg, commit `93f089a333a43089fb1a08a73c37d05fd6683214`
+
+## When To Consider
+
+MaxwellLink is a modular Python framework for self-consistent electromagnetic molecular co-simulation, coupling EM solvers with quantum, force-field, and first-principles molecular dynamics drivers.
+
+## DeepScientist Runtime Rule
+
+This card is package knowledge and routing context only. It does not mean the
+solver, Python module, CLI binary, compiled backend, license server, dataset, or
+HPC module is installed in the active environment. Before computed work, use
+`bash_exec(...)` to perform an import, executable, version, and smoke-test check
+appropriate for `maxwelllink`.
+
+## Package Check
+
+For CLI/HPC-oriented environments, check the executable or loaded module before
+running any expensive job:
+
+```bash
+command -v maxwelllink || true
+maxwelllink --version || true
+```
+
+If the package is available only through environment modules, record the module
+state and the exact executable path in `validation/environment/maxwelllink_doctor.json`.
+
+Record the result with `artifact.science(...)` as `science.package_check`. Use
+`status="passed"` only when the environment can run at least a minimal smoke
+path. Use `status="failed"` or `status="blocked"` when the check explains why
+execution cannot proceed.
+
+Generated import or executable names are starting points. If `maxwelllink` uses
+a different Python module, CLI binary, environment module, container, or wrapper
+script, adjust the check before concluding the solver is unavailable.
+
+## Expected Science Nodes
+
+- `science.package_check` for import/executable/version/smoke-test evidence
+- `science.computational_run` for solver execution, simulation, fitting, or numerical computation
+- `science.dataset_analysis` when the task primarily analyzes existing data
+- `science.parameter_sweep` when varying parameters, inputs, models, or solver settings
+- `science.validation_result` for convergence, units, schema, controls, or correctness checks
+- `science.claim` only after evidence paths or related nodes support the claim
+
+## Evidence Path Conventions
+
+- `simulations/inputs/` for generated or selected solver inputs
+- `simulations/logs/` for stdout, stderr, scheduler logs, or solver logs
+- `simulations/outputs/` for structured run outputs
+- `analyses/scripts/`, `analyses/logs/`, and `analyses/outputs/` for dataset analysis
+- `validation/environment/` for package checks
+- `validation/runs/` for convergence, unit, schema, or correctness sidecars
+- `figures/` for derived visualizations
+
+## Validation Checklist
+
+- Record package version, executable path, backend, module state, or container image when relevant.
+- Preserve input files and parameters that define the scientific state.
+- Capture units, coordinate conventions, timestep/mesh/basis/model settings, seeds, and convergence criteria when applicable.
+- Validate output schema and important physical or statistical invariants before recording a computed claim.
+- Link claims to run, analysis, sweep, and validation nodes rather than relying on prose.
+
+## Common Pitfalls
+
+- Do not treat the package card or knowledge URL as runtime availability.
+- Do not weaken solver tolerances, physical models, dataset filters, or convergence criteria to make a run pass unless the change is explicitly part of the scientific question.
+- Do not call a value `computed` unless the corresponding run or analysis happened in the current quest and evidence paths are recorded.
+- Do not copy package knowledge-base material into the quest without preserving its source and license context.

package/src/skills/science/references/packages/mcdc.md

@@ -0,0 +1,73 @@
+# MC/DC Neutron Transport
+
+## Catalog
+
+- Package id: `mcdc`
+- Domains: `high_energy_physics`
+- Tags: `neutron-transport`, `monte-carlo`, `reactor-physics`, `radiation-transport`, `computational-physics`
+- Knowledge URL: https://github.com/skilled-scipkg/MCDC
+- Source archive URL: https://github.com/skilled-scipkg/MCDC/archive/refs/heads/dev.zip
+- Upstream project URL: https://github.com/CEMeNT-PSAAP/MCDC
+- Homepage: https://mcdc.readthedocs.io/en/latest/
+- Catalog source: FermiLink skilled-scipkg, commit `93f089a333a43089fb1a08a73c37d05fd6683214`
+
+## When To Consider
+
+MC/DC is a scalable Monte Carlo neutron transport package for fully transient reactor and radiation simulations, enabling rapid methods development with performant, machine-portable Python implementations.
+
+## DeepScientist Runtime Rule
+
+This card is package knowledge and routing context only. It does not mean the
+solver, Python module, CLI binary, compiled backend, license server, dataset, or
+HPC module is installed in the active environment. Before computed work, use
+`bash_exec(...)` to perform an import, executable, version, and smoke-test check
+appropriate for `mcdc`.
+
+## Package Check
+
+Use a package-specific import, executable, or module check and save the result
+under `validation/environment/mcdc_doctor.json` before treating the
+runtime as usable.
+
+Record the result with `artifact.science(...)` as `science.package_check`. Use
+`status="passed"` only when the environment can run at least a minimal smoke
+path. Use `status="failed"` or `status="blocked"` when the check explains why
+execution cannot proceed.
+
+Generated import or executable names are starting points. If `mcdc` uses
+a different Python module, CLI binary, environment module, container, or wrapper
+script, adjust the check before concluding the solver is unavailable.
+
+## Expected Science Nodes
+
+- `science.package_check` for import/executable/version/smoke-test evidence
+- `science.computational_run` for solver execution, simulation, fitting, or numerical computation
+- `science.dataset_analysis` when the task primarily analyzes existing data
+- `science.parameter_sweep` when varying parameters, inputs, models, or solver settings
+- `science.validation_result` for convergence, units, schema, controls, or correctness checks
+- `science.claim` only after evidence paths or related nodes support the claim
+
+## Evidence Path Conventions
+
+- `simulations/inputs/` for generated or selected solver inputs
+- `simulations/logs/` for stdout, stderr, scheduler logs, or solver logs
+- `simulations/outputs/` for structured run outputs
+- `analyses/scripts/`, `analyses/logs/`, and `analyses/outputs/` for dataset analysis
+- `validation/environment/` for package checks
+- `validation/runs/` for convergence, unit, schema, or correctness sidecars
+- `figures/` for derived visualizations
+
+## Validation Checklist
+
+- Record package version, executable path, backend, module state, or container image when relevant.
+- Preserve input files and parameters that define the scientific state.
+- Capture units, coordinate conventions, timestep/mesh/basis/model settings, seeds, and convergence criteria when applicable.
+- Validate output schema and important physical or statistical invariants before recording a computed claim.
+- Link claims to run, analysis, sweep, and validation nodes rather than relying on prose.
+
+## Common Pitfalls
+
+- Do not treat the package card or knowledge URL as runtime availability.
+- Do not weaken solver tolerances, physical models, dataset filters, or convergence criteria to make a run pass unless the change is explicitly part of the scientific question.
+- Do not call a value `computed` unless the corresponding run or analysis happened in the current quest and evidence paths are recorded.
+- Do not copy package knowledge-base material into the quest without preserving its source and license context.

package/src/skills/science/references/packages/meep.md

@@ -0,0 +1,80 @@
+# Meep Electromagnetic FDTD Simulator
+
+## Catalog
+
+- Package id: `meep`
+- Domains: `electromagnetics`
+- Tags: `electromagnetics`, `fdtd`, `photonics`, `nanophotonics`, `maxwell`, `wave-optics`
+- Knowledge URL: https://github.com/skilled-scipkg/meep
+- Source archive URL: https://github.com/skilled-scipkg/meep/archive/refs/heads/master.zip
+- Upstream project URL: https://github.com/NanoComp/meep
+- Homepage: https://github.com/NanoComp/meep
+- Catalog source: FermiLink skilled-scipkg, commit `93f089a333a43089fb1a08a73c37d05fd6683214`
+
+## When To Consider
+
+Meep is an open-source finite-difference time-domain simulator for modeling electromagnetic wave propagation, resonances, and scattering in photonic structures across 1D, 2D, 3D, and cylindrical geometries.
+
+## DeepScientist Runtime Rule
+
+This card is package knowledge and routing context only. It does not mean the
+solver, Python module, CLI binary, compiled backend, license server, dataset, or
+HPC module is installed in the active environment. Before computed work, use
+`bash_exec(...)` to perform an import, executable, version, and smoke-test check
+appropriate for `meep`.
+
+## Package Check
+
+For CLI/HPC-oriented environments, check the executable or loaded module before
+running any expensive job:
+
+```bash
+command -v meep || true
+meep --version || true
+```
+
+If the package is available only through environment modules, record the module
+state and the exact executable path in `validation/environment/meep_doctor.json`.
+
+Record the result with `artifact.science(...)` as `science.package_check`. Use
+`status="passed"` only when the environment can run at least a minimal smoke
+path. Use `status="failed"` or `status="blocked"` when the check explains why
+execution cannot proceed.
+
+Generated import or executable names are starting points. If `meep` uses
+a different Python module, CLI binary, environment module, container, or wrapper
+script, adjust the check before concluding the solver is unavailable.
+
+## Expected Science Nodes
+
+- `science.package_check` for import/executable/version/smoke-test evidence
+- `science.computational_run` for solver execution, simulation, fitting, or numerical computation
+- `science.dataset_analysis` when the task primarily analyzes existing data
+- `science.parameter_sweep` when varying parameters, inputs, models, or solver settings
+- `science.validation_result` for convergence, units, schema, controls, or correctness checks
+- `science.claim` only after evidence paths or related nodes support the claim
+
+## Evidence Path Conventions
+
+- `simulations/inputs/` for generated or selected solver inputs
+- `simulations/logs/` for stdout, stderr, scheduler logs, or solver logs
+- `simulations/outputs/` for structured run outputs
+- `analyses/scripts/`, `analyses/logs/`, and `analyses/outputs/` for dataset analysis
+- `validation/environment/` for package checks
+- `validation/runs/` for convergence, unit, schema, or correctness sidecars
+- `figures/` for derived visualizations
+
+## Validation Checklist
+
+- Record package version, executable path, backend, module state, or container image when relevant.
+- Preserve input files and parameters that define the scientific state.
+- Capture units, coordinate conventions, timestep/mesh/basis/model settings, seeds, and convergence criteria when applicable.
+- Validate output schema and important physical or statistical invariants before recording a computed claim.
+- Link claims to run, analysis, sweep, and validation nodes rather than relying on prose.
+
+## Common Pitfalls
+
+- Do not treat the package card or knowledge URL as runtime availability.
+- Do not weaken solver tolerances, physical models, dataset filters, or convergence criteria to make a run pass unless the change is explicitly part of the scientific question.
+- Do not call a value `computed` unless the corresponding run or analysis happened in the current quest and evidence paths are recorded.
+- Do not copy package knowledge-base material into the quest without preserving its source and license context.

package/src/skills/science/references/packages/mfem.md

@@ -0,0 +1,80 @@
+# MFEM Finite Element Library
+
+## Catalog
+
+- Package id: `mfem`
+- Domains: `computational_fluid_dynamics`, `finite_element_engineering`, `workflow_provenance`
+- Tags: `finite-element`, `pde`, `scientific-computing`, `hpc`, `cpp`, `amr`
+- Knowledge URL: https://github.com/skilled-scipkg/mfem
+- Source archive URL: https://github.com/skilled-scipkg/mfem/archive/refs/heads/master.zip
+- Upstream project URL: https://github.com/mfem/mfem
+- Homepage: http://mfem.org
+- Catalog source: FermiLink skilled-scipkg, commit `93f089a333a43089fb1a08a73c37d05fd6683214`
+
+## When To Consider
+
+MFEM is a modular, scalable C++ finite element library for high-order PDE discretization, adaptive meshing, and MPI/GPU-enabled scientific simulation from laptops to supercomputers.
+
+## DeepScientist Runtime Rule
+
+This card is package knowledge and routing context only. It does not mean the
+solver, Python module, CLI binary, compiled backend, license server, dataset, or
+HPC module is installed in the active environment. Before computed work, use
+`bash_exec(...)` to perform an import, executable, version, and smoke-test check
+appropriate for `mfem`.
+
+## Package Check
+
+For CLI/HPC-oriented environments, check the executable or loaded module before
+running any expensive job:
+
+```bash
+command -v mfem || true
+mfem --version || true
+```
+
+If the package is available only through environment modules, record the module
+state and the exact executable path in `validation/environment/mfem_doctor.json`.
+
+Record the result with `artifact.science(...)` as `science.package_check`. Use
+`status="passed"` only when the environment can run at least a minimal smoke
+path. Use `status="failed"` or `status="blocked"` when the check explains why
+execution cannot proceed.
+
+Generated import or executable names are starting points. If `mfem` uses
+a different Python module, CLI binary, environment module, container, or wrapper
+script, adjust the check before concluding the solver is unavailable.
+
+## Expected Science Nodes
+
+- `science.package_check` for import/executable/version/smoke-test evidence
+- `science.computational_run` for solver execution, simulation, fitting, or numerical computation
+- `science.dataset_analysis` when the task primarily analyzes existing data
+- `science.parameter_sweep` when varying parameters, inputs, models, or solver settings
+- `science.validation_result` for convergence, units, schema, controls, or correctness checks
+- `science.claim` only after evidence paths or related nodes support the claim
+
+## Evidence Path Conventions
+
+- `simulations/inputs/` for generated or selected solver inputs
+- `simulations/logs/` for stdout, stderr, scheduler logs, or solver logs
+- `simulations/outputs/` for structured run outputs
+- `analyses/scripts/`, `analyses/logs/`, and `analyses/outputs/` for dataset analysis
+- `validation/environment/` for package checks
+- `validation/runs/` for convergence, unit, schema, or correctness sidecars
+- `figures/` for derived visualizations
+
+## Validation Checklist
+
+- Record package version, executable path, backend, module state, or container image when relevant.
+- Preserve input files and parameters that define the scientific state.
+- Capture units, coordinate conventions, timestep/mesh/basis/model settings, seeds, and convergence criteria when applicable.
+- Validate output schema and important physical or statistical invariants before recording a computed claim.
+- Link claims to run, analysis, sweep, and validation nodes rather than relying on prose.
+
+## Common Pitfalls
+
+- Do not treat the package card or knowledge URL as runtime availability.
+- Do not weaken solver tolerances, physical models, dataset filters, or convergence criteria to make a run pass unless the change is explicitly part of the scientific question.
+- Do not call a value `computed` unless the corresponding run or analysis happened in the current quest and evidence paths are recorded.
+- Do not copy package knowledge-base material into the quest without preserving its source and license context.

package/src/skills/science/references/packages/mitgcm.md

@@ -0,0 +1,73 @@
+# MIT General Circulation Model (MITgcm)
+
+## Catalog
+
+- Package id: `mitgcm`
+- Domains: `computational_fluid_dynamics`
+- Tags: `mitgcm`, `ocean-model`, `climate-model`, `atmospheric-circulation`, `geophysical-fluids`, `sea-ice`
+- Knowledge URL: https://github.com/skilled-scipkg/MITgcm
+- Source archive URL: https://github.com/skilled-scipkg/MITgcm/archive/refs/heads/master.zip
+- Upstream project URL: https://github.com/MITgcm/MITgcm
+- Homepage: http://mitgcm.org/
+- Catalog source: FermiLink skilled-scipkg, commit `93f089a333a43089fb1a08a73c37d05fd6683214`
+
+## When To Consider
+
+MITgcm is a flexible geophysical fluid dynamics package for simulating ocean and atmospheric circulation, with coupled ocean sea ice and climate process capabilities.
+
+## DeepScientist Runtime Rule
+
+This card is package knowledge and routing context only. It does not mean the
+solver, Python module, CLI binary, compiled backend, license server, dataset, or
+HPC module is installed in the active environment. Before computed work, use
+`bash_exec(...)` to perform an import, executable, version, and smoke-test check
+appropriate for `mitgcm`.
+
+## Package Check
+
+Use a package-specific import, executable, or module check and save the result
+under `validation/environment/mitgcm_doctor.json` before treating the
+runtime as usable.
+
+Record the result with `artifact.science(...)` as `science.package_check`. Use
+`status="passed"` only when the environment can run at least a minimal smoke
+path. Use `status="failed"` or `status="blocked"` when the check explains why
+execution cannot proceed.
+
+Generated import or executable names are starting points. If `mitgcm` uses
+a different Python module, CLI binary, environment module, container, or wrapper
+script, adjust the check before concluding the solver is unavailable.
+
+## Expected Science Nodes
+
+- `science.package_check` for import/executable/version/smoke-test evidence
+- `science.computational_run` for solver execution, simulation, fitting, or numerical computation
+- `science.dataset_analysis` when the task primarily analyzes existing data
+- `science.parameter_sweep` when varying parameters, inputs, models, or solver settings
+- `science.validation_result` for convergence, units, schema, controls, or correctness checks
+- `science.claim` only after evidence paths or related nodes support the claim
+
+## Evidence Path Conventions
+
+- `simulations/inputs/` for generated or selected solver inputs
+- `simulations/logs/` for stdout, stderr, scheduler logs, or solver logs
+- `simulations/outputs/` for structured run outputs
+- `analyses/scripts/`, `analyses/logs/`, and `analyses/outputs/` for dataset analysis
+- `validation/environment/` for package checks
+- `validation/runs/` for convergence, unit, schema, or correctness sidecars
+- `figures/` for derived visualizations
+
+## Validation Checklist
+
+- Record package version, executable path, backend, module state, or container image when relevant.
+- Preserve input files and parameters that define the scientific state.
+- Capture units, coordinate conventions, timestep/mesh/basis/model settings, seeds, and convergence criteria when applicable.
+- Validate output schema and important physical or statistical invariants before recording a computed claim.
+- Link claims to run, analysis, sweep, and validation nodes rather than relying on prose.
+
+## Common Pitfalls
+
+- Do not treat the package card or knowledge URL as runtime availability.
+- Do not weaken solver tolerances, physical models, dataset filters, or convergence criteria to make a run pass unless the change is explicitly part of the scientific question.
+- Do not call a value `computed` unless the corresponding run or analysis happened in the current quest and evidence paths are recorded.
+- Do not copy package knowledge-base material into the quest without preserving its source and license context.

package/src/skills/science/references/packages/modflow6.md

@@ -0,0 +1,73 @@
+# MODFLOW 6 Groundwater Model
+
+## Catalog
+
+- Package id: `modflow6`
+- Domains: `computational_science`
+- Tags: `groundwater`, `hydrology`, `hydrogeology`, `usgs`, `modflow`, `aquifer-modeling`
+- Knowledge URL: https://github.com/skilled-scipkg/modflow6
+- Source archive URL: https://github.com/skilled-scipkg/modflow6/archive/refs/heads/develop.zip
+- Upstream project URL: https://github.com/MODFLOW-ORG/modflow6
+- Homepage: https://modflow6.readthedocs.io/
+- Catalog source: FermiLink skilled-scipkg, commit `93f089a333a43089fb1a08a73c37d05fd6683214`
+
+## When To Consider
+
+MODFLOW 6 is the USGS modular hydrologic simulator for groundwater flow and transport, supporting integrated surface-water interactions, time-varying stresses, and extensible packages for basin-scale hydrogeologic modeling.
+
+## DeepScientist Runtime Rule
+
+This card is package knowledge and routing context only. It does not mean the
+solver, Python module, CLI binary, compiled backend, license server, dataset, or
+HPC module is installed in the active environment. Before computed work, use
+`bash_exec(...)` to perform an import, executable, version, and smoke-test check
+appropriate for `modflow6`.
+
+## Package Check
+
+Use a package-specific import, executable, or module check and save the result
+under `validation/environment/modflow6_doctor.json` before treating the
+runtime as usable.
+
+Record the result with `artifact.science(...)` as `science.package_check`. Use
+`status="passed"` only when the environment can run at least a minimal smoke
+path. Use `status="failed"` or `status="blocked"` when the check explains why
+execution cannot proceed.
+
+Generated import or executable names are starting points. If `modflow6` uses
+a different Python module, CLI binary, environment module, container, or wrapper
+script, adjust the check before concluding the solver is unavailable.
+
+## Expected Science Nodes
+
+- `science.package_check` for import/executable/version/smoke-test evidence
+- `science.computational_run` for solver execution, simulation, fitting, or numerical computation
+- `science.dataset_analysis` when the task primarily analyzes existing data
+- `science.parameter_sweep` when varying parameters, inputs, models, or solver settings
+- `science.validation_result` for convergence, units, schema, controls, or correctness checks
+- `science.claim` only after evidence paths or related nodes support the claim
+
+## Evidence Path Conventions
+
+- `simulations/inputs/` for generated or selected solver inputs
+- `simulations/logs/` for stdout, stderr, scheduler logs, or solver logs
+- `simulations/outputs/` for structured run outputs
+- `analyses/scripts/`, `analyses/logs/`, and `analyses/outputs/` for dataset analysis
+- `validation/environment/` for package checks
+- `validation/runs/` for convergence, unit, schema, or correctness sidecars
+- `figures/` for derived visualizations
+
+## Validation Checklist
+
+- Record package version, executable path, backend, module state, or container image when relevant.
+- Preserve input files and parameters that define the scientific state.
+- Capture units, coordinate conventions, timestep/mesh/basis/model settings, seeds, and convergence criteria when applicable.
+- Validate output schema and important physical or statistical invariants before recording a computed claim.
+- Link claims to run, analysis, sweep, and validation nodes rather than relying on prose.
+
+## Common Pitfalls
+
+- Do not treat the package card or knowledge URL as runtime availability.
+- Do not weaken solver tolerances, physical models, dataset filters, or convergence criteria to make a run pass unless the change is explicitly part of the scientific question.
+- Do not call a value `computed` unless the corresponding run or analysis happened in the current quest and evidence paths are recorded.
+- Do not copy package knowledge-base material into the quest without preserving its source and license context.

package/src/skills/science/references/packages/molecool.md

@@ -0,0 +1,73 @@
+# MoleCool Molecular Laser Cooling Simulator
+
+## Catalog
+
+- Package id: `molecool`
+- Domains: `molecular_dynamics`, `high_energy_physics`
+- Tags: `amo-physics`, `laser-cooling`, `molecular-spectroscopy`, `optical-bloch`, `rate-equations`, `monte-carlo`
+- Knowledge URL: https://github.com/skilled-scipkg/MoleCool
+- Source archive URL: https://github.com/skilled-scipkg/MoleCool/archive/refs/heads/master.zip
+- Upstream project URL: https://github.com/LangenGroup/MoleCool
+- Homepage: https://arxiv.org/abs/2510.16203
+- Catalog source: FermiLink skilled-scipkg, commit `93f089a333a43089fb1a08a73c37d05fd6683214`
+
+## When To Consider
+
+MoleCool simulates laser-driven molecular dynamics and spectra, combining optical Bloch and rate-equation solvers, cooling-force mapping, and Monte Carlo trajectories for AMO cooling and trapping experiment design.
+
+## DeepScientist Runtime Rule
+
+This card is package knowledge and routing context only. It does not mean the
+solver, Python module, CLI binary, compiled backend, license server, dataset, or
+HPC module is installed in the active environment. Before computed work, use
+`bash_exec(...)` to perform an import, executable, version, and smoke-test check
+appropriate for `molecool`.
+
+## Package Check
+
+Use a package-specific import, executable, or module check and save the result
+under `validation/environment/molecool_doctor.json` before treating the
+runtime as usable.
+
+Record the result with `artifact.science(...)` as `science.package_check`. Use
+`status="passed"` only when the environment can run at least a minimal smoke
+path. Use `status="failed"` or `status="blocked"` when the check explains why
+execution cannot proceed.
+
+Generated import or executable names are starting points. If `molecool` uses
+a different Python module, CLI binary, environment module, container, or wrapper
+script, adjust the check before concluding the solver is unavailable.
+
+## Expected Science Nodes
+
+- `science.package_check` for import/executable/version/smoke-test evidence
+- `science.computational_run` for solver execution, simulation, fitting, or numerical computation
+- `science.dataset_analysis` when the task primarily analyzes existing data
+- `science.parameter_sweep` when varying parameters, inputs, models, or solver settings
+- `science.validation_result` for convergence, units, schema, controls, or correctness checks
+- `science.claim` only after evidence paths or related nodes support the claim
+
+## Evidence Path Conventions
+
+- `simulations/inputs/` for generated or selected solver inputs
+- `simulations/logs/` for stdout, stderr, scheduler logs, or solver logs
+- `simulations/outputs/` for structured run outputs
+- `analyses/scripts/`, `analyses/logs/`, and `analyses/outputs/` for dataset analysis
+- `validation/environment/` for package checks
+- `validation/runs/` for convergence, unit, schema, or correctness sidecars
+- `figures/` for derived visualizations
+
+## Validation Checklist
+
+- Record package version, executable path, backend, module state, or container image when relevant.
+- Preserve input files and parameters that define the scientific state.
+- Capture units, coordinate conventions, timestep/mesh/basis/model settings, seeds, and convergence criteria when applicable.
+- Validate output schema and important physical or statistical invariants before recording a computed claim.
+- Link claims to run, analysis, sweep, and validation nodes rather than relying on prose.
+
+## Common Pitfalls
+
+- Do not treat the package card or knowledge URL as runtime availability.
+- Do not weaken solver tolerances, physical models, dataset filters, or convergence criteria to make a run pass unless the change is explicitly part of the scientific question.
+- Do not call a value `computed` unless the corresponding run or analysis happened in the current quest and evidence paths are recorded.
+- Do not copy package knowledge-base material into the quest without preserving its source and license context.

package/src/skills/science/references/packages/mom6.md

@@ -0,0 +1,73 @@
+# MOM6 Modular Ocean Model
+
+## Catalog
+
+- Package id: `mom6`
+- Domains: `computational_science`
+- Tags: `ocean-modeling`, `climate-modeling`, `ocean-circulation`, `geophysical-fluid-dynamics`, `earth-system`
+- Knowledge URL: https://github.com/skilled-scipkg/MOM6
+- Source archive URL: https://github.com/skilled-scipkg/MOM6/archive/refs/heads/main.zip
+- Upstream project URL: https://github.com/mom-ocean/MOM6
+- Homepage: https://github.com/mom-ocean/MOM6
+- Catalog source: FermiLink skilled-scipkg, commit `93f089a333a43089fb1a08a73c37d05fd6683214`
+
+## When To Consider
+
+Modular Ocean Model (MOM6) is a finite-volume ocean general circulation model for ocean-only and coupled climate simulations, supporting tracer transport, stratified dynamics, and realistic large-scale ocean process studies.
+
+## DeepScientist Runtime Rule
+
+This card is package knowledge and routing context only. It does not mean the
+solver, Python module, CLI binary, compiled backend, license server, dataset, or
+HPC module is installed in the active environment. Before computed work, use
+`bash_exec(...)` to perform an import, executable, version, and smoke-test check
+appropriate for `mom6`.
+
+## Package Check
+
+Use a package-specific import, executable, or module check and save the result
+under `validation/environment/mom6_doctor.json` before treating the
+runtime as usable.
+
+Record the result with `artifact.science(...)` as `science.package_check`. Use
+`status="passed"` only when the environment can run at least a minimal smoke
+path. Use `status="failed"` or `status="blocked"` when the check explains why
+execution cannot proceed.
+
+Generated import or executable names are starting points. If `mom6` uses
+a different Python module, CLI binary, environment module, container, or wrapper
+script, adjust the check before concluding the solver is unavailable.
+
+## Expected Science Nodes
+
+- `science.package_check` for import/executable/version/smoke-test evidence
+- `science.computational_run` for solver execution, simulation, fitting, or numerical computation
+- `science.dataset_analysis` when the task primarily analyzes existing data
+- `science.parameter_sweep` when varying parameters, inputs, models, or solver settings
+- `science.validation_result` for convergence, units, schema, controls, or correctness checks
+- `science.claim` only after evidence paths or related nodes support the claim
+
+## Evidence Path Conventions
+
+- `simulations/inputs/` for generated or selected solver inputs
+- `simulations/logs/` for stdout, stderr, scheduler logs, or solver logs
+- `simulations/outputs/` for structured run outputs
+- `analyses/scripts/`, `analyses/logs/`, and `analyses/outputs/` for dataset analysis
+- `validation/environment/` for package checks
+- `validation/runs/` for convergence, unit, schema, or correctness sidecars
+- `figures/` for derived visualizations
+
+## Validation Checklist
+
+- Record package version, executable path, backend, module state, or container image when relevant.
+- Preserve input files and parameters that define the scientific state.
+- Capture units, coordinate conventions, timestep/mesh/basis/model settings, seeds, and convergence criteria when applicable.
+- Validate output schema and important physical or statistical invariants before recording a computed claim.
+- Link claims to run, analysis, sweep, and validation nodes rather than relying on prose.
+
+## Common Pitfalls
+
+- Do not treat the package card or knowledge URL as runtime availability.
+- Do not weaken solver tolerances, physical models, dataset filters, or convergence criteria to make a run pass unless the change is explicitly part of the scientific question.
+- Do not call a value `computed` unless the corresponding run or analysis happened in the current quest and evidence paths are recorded.
+- Do not copy package knowledge-base material into the quest without preserving its source and license context.