@researai/deepscientist 1.5.16 → 1.6.0

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

@@ -0,0 +1,73 @@
+# Thermo Chemical Engineering Thermodynamics
+
+## Catalog
+
+- Package id: `thermo`
+- Domains: `workflow_provenance`
+- Tags: `thermodynamics`, `chemicalengineering`, `phaseequilibrium`, `vle`, `flashcalculation`, `mixtureproperties`
+- Knowledge URL: https://github.com/skilled-scipkg/thermo
+- Source archive URL: https://github.com/skilled-scipkg/thermo/archive/refs/heads/master.zip
+- Upstream project URL: https://github.com/CalebBell/thermo
+- Homepage: https://github.com/CalebBell/thermo
+- Catalog source: FermiLink skilled-scipkg, commit `93f089a333a43089fb1a08a73c37d05fd6683214`
+
+## When To Consider
+
+Python package for chemical engineering thermodynamics and phase equilibrium, providing pure component and mixture property models, transport correlations, and rigorous flash calculation workflows.
+
+## 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 `thermo`.
+
+## Package Check
+
+Use a package-specific import, executable, or module check and save the result
+under `validation/environment/thermo_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 `thermo` 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/tkwant.md

@@ -0,0 +1,73 @@
+# Tkwant Time-Dependent Quantum Transport
+
+## Catalog
+
+- Package id: `tkwant`
+- Domains: `quantum_chemistry`, `plasma_particle_simulation`
+- Tags: `time-dependent-quantum-transport`, `non-equilibrium`, `tight-binding`, `mesoscopic-physics`, `nanoelectronics`, `keldysh-formalism`
+- Knowledge URL: https://github.com/skilled-scipkg/tkwant
+- Source archive URL: https://github.com/skilled-scipkg/tkwant/archive/refs/heads/master.zip
+- Upstream project URL: https://gitlab.kwant-project.org/kwant/tkwant
+- Homepage: https://tkwant.kwant-project.org/
+- Catalog source: FermiLink skilled-scipkg, commit `93f089a333a43089fb1a08a73c37d05fd6683214`
+
+## When To Consider
+
+Tkwant is a Python package for simulating time-dependent quantum dynamics of mesoscopic systems and serves as the time-dependent extension of Kwant for open quantum transport problems.
+
+## 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 `tkwant`.
+
+## Package Check
+
+Use a package-specific import, executable, or module check and save the result
+under `validation/environment/tkwant_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 `tkwant` 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/tvb-root.md

@@ -0,0 +1,73 @@
+# The Virtual Brain Core
+
+## Catalog
+
+- Package id: `tvb-root`
+- Domains: `bioinformatics`, `computational_neuroscience`, `workflow_provenance`
+- Tags: `neuroscience`, `brain-simulation`, `connectomics`, `neuroinformatics`, `computational-neuroscience`
+- Knowledge URL: https://github.com/skilled-scipkg/tvb-root
+- Source archive URL: https://github.com/skilled-scipkg/tvb-root/archive/refs/heads/master.zip
+- Upstream project URL: https://github.com/the-virtual-brain/tvb-root
+- Homepage: https://thevirtualbrain.org
+- Catalog source: FermiLink skilled-scipkg, commit `93f089a333a43089fb1a08a73c37d05fd6683214`
+
+## When To Consider
+
+The Virtual Brain main codebase provides neuroscience tools for whole-brain network simulation, multimodal data analysis, and framework components for workflows, storage, and visualization.
+
+## 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 `tvb-root`.
+
+## Package Check
+
+Use a package-specific import, executable, or module check and save the result
+under `validation/environment/tvb-root_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 `tvb-root` 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/uproot5.md

@@ -0,0 +1,73 @@
+# Uproot5 ROOT I/O
+
+## Catalog
+
+- Package id: `uproot5`
+- Domains: `high_energy_physics`, `workflow_provenance`
+- Tags: `hep`, `root`, `ttree`, `columnar`, `particle-physics`
+- Knowledge URL: https://github.com/skilled-scipkg/uproot5
+- Source archive URL: https://github.com/skilled-scipkg/uproot5/archive/refs/heads/main.zip
+- Upstream project URL: https://github.com/scikit-hep/uproot5
+- Homepage: https://uproot.readthedocs.io
+- Catalog source: FermiLink skilled-scipkg, commit `93f089a333a43089fb1a08a73c37d05fd6683214`
+
+## When To Consider
+
+Uproot5 is a pure Python library for reading and writing CERN ROOT files, especially TTrees, and streaming high-energy physics event data into array and dataframe workflows without C++ ROOT.
+
+## 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 `uproot5`.
+
+## Package Check
+
+Use a package-specific import, executable, or module check and save the result
+under `validation/environment/uproot5_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 `uproot5` 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/vampire.md

@@ -0,0 +1,80 @@
+# VAMPIRE Atomistic Spin Dynamics Simulator
+
+## Catalog
+
+- Package id: `vampire`
+- Domains: `molecular_dynamics`, `materials_science`
+- Tags: `magnetism`, `spin-dynamics`, `atomistic-simulation`, `materials-science`, `computational-physics`
+- Knowledge URL: https://github.com/skilled-scipkg/vampire
+- Source archive URL: https://github.com/skilled-scipkg/vampire/archive/refs/heads/main.zip
+- Upstream project URL: https://github.com/richard-evans/vampire
+- Homepage: https://github.com/richard-evans/vampire
+- Catalog source: FermiLink skilled-scipkg, commit `93f089a333a43089fb1a08a73c37d05fd6683214`
+
+## When To Consider
+
+VAMPIRE is an atomistic spin dynamics code for simulating magnetic materials, enabling finite-temperature magnetism, hysteresis, ultrafast dynamics, and spin-transport studies at atomic resolution.
+
+## 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 `vampire`.
+
+## Package Check
+
+For CLI/HPC-oriented environments, check the executable or loaded module before
+running any expensive job:
+
+```bash
+command -v vampire || true
+vampire --version || true
+```
+
+If the package is available only through environment modules, record the module
+state and the exact executable path in `validation/environment/vampire_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 `vampire` 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/wannier_tools.md

@@ -0,0 +1,73 @@
+# WannierTools Topological Materials Toolkit
+
+## Catalog
+
+- Package id: `wannier_tools`
+- Domains: `quantum_chemistry`, `finite_element_engineering`
+- Tags: `topological-materials`, `tight-binding`, `electronic-structure`, `wannier`, `condensed-matter`, `surface-states`
+- Knowledge URL: https://github.com/skilled-scipkg/wannier_tools
+- Source archive URL: https://github.com/skilled-scipkg/wannier_tools/archive/refs/heads/master.zip
+- Upstream project URL: https://github.com/quanshengwu/wannier_tools
+- Homepage: http://www.wanniertools.org
+- Catalog source: FermiLink skilled-scipkg, commit `93f089a333a43089fb1a08a73c37d05fd6683214`
+
+## When To Consider
+
+WannierTools analyzes Wannier tight binding Hamiltonians to compute topological invariants, surface spectra, Weyl and Dirac features, Berry properties, and transport observables in quantum materials.
+
+## 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 `wannier_tools`.
+
+## Package Check
+
+Use a package-specific import, executable, or module check and save the result
+under `validation/environment/wannier_tools_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 `wannier_tools` 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/warpx.md

@@ -0,0 +1,80 @@
+# WarpX PIC Plasma Simulator
+
+## Catalog
+
+- Package id: `warpx`
+- Domains: `electromagnetics`, `plasma_particle_simulation`, `workflow_provenance`
+- Tags: `plasma-physics`, `particle-in-cell`, `accelerator-physics`, `electromagnetics`, `electrostatics`, `hpc-simulation`
+- Knowledge URL: https://github.com/skilled-scipkg/warpx
+- Source archive URL: https://github.com/skilled-scipkg/warpx/archive/refs/heads/development.zip
+- Upstream project URL: https://github.com/BLAST-WarpX/warpx
+- Homepage: https://blast-warpx.github.io
+- Catalog source: FermiLink skilled-scipkg, commit `93f089a333a43089fb1a08a73c37d05fd6683214`
+
+## When To Consider
+
+WarpX is a high-performance electromagnetic and electrostatic particle-in-cell simulator for plasma and accelerator modeling, with mesh refinement, boosted-frame techniques, and scalable CPU and GPU execution on 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 `warpx`.
+
+## Package Check
+
+For CLI/HPC-oriented environments, check the executable or loaded module before
+running any expensive job:
+
+```bash
+command -v warpx || true
+warpx --version || true
+```
+
+If the package is available only through environment modules, record the module
+state and the exact executable path in `validation/environment/warpx_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 `warpx` 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/wrf.md

@@ -0,0 +1,73 @@
+# WRF Weather Forecasting Model
+
+## Catalog
+
+- Package id: `wrf`
+- Domains: `computational_science`
+- Tags: `wrf`, `atmospheric-modeling`, `numerical-weather-prediction`, `mesoscale`, `regional-climate`, `weather-forecasting`
+- Knowledge URL: https://github.com/skilled-scipkg/WRF
+- Source archive URL: https://github.com/skilled-scipkg/WRF/archive/refs/heads/master.zip
+- Upstream project URL: https://github.com/wrf-model/WRF
+- Homepage: https://github.com/wrf-model/WRF
+- Catalog source: FermiLink skilled-scipkg, commit `93f089a333a43089fb1a08a73c37d05fd6683214`
+
+## When To Consider
+
+WRF is an open source mesoscale numerical weather prediction system for atmospheric research and operational forecasting, including regional climate downscaling, severe weather studies, and nested domain simulations.
+
+## 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 `wrf`.
+
+## Package Check
+
+Use a package-specific import, executable, or module check and save the result
+under `validation/environment/wrf_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 `wrf` 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.