Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Inorganic chemistry, physical chemistry, and materials science — crystal structures, coordination chemistry, lattice parameters, thermodynamic properties, electronic structure. Use for unit cell volume calculations, coordination geometry, materials property estimation, and inorganic-mechanism reasoning. Complementary to tooluniverse-organic-chemistry.
.claude/skills/tooluniverse-inorganic-physical-chemistry/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-06 | ✗→✓ | ▲ Improved | 122% | 0% |
| case-13 | ✗→✓ | ▲ Improved | 202% | 0% |
| case-16 | ✓→✓ | = Same ✓ | 9% | 0% |
| case-01 | ✗→✗ | = Same ✗ | 109% | 0% |
| case-02 | ✗→✗ | = Same ✗ | 111% | 0% |
When given crystal structure data, always COMPUTE don't guess:
CrystalStructure_validate tool (via MCP/SDK). Fallback: python3 skills/tooluniverse-organic-chemistry/scripts/crystal_validator.py --a X --b Y --c Z --alpha A --beta B --gamma G --Z N --MW M --density D--datasets path/to/datasets.jsonKey reasoning patterns:
python3 skills/tooluniverse-organic-chemistry/scripts/chemistry_facts.py point_groups for point group lookupCOMPUTE DON'T ESTIMATE — write Python code for:
Preferred: Use EquilibriumSolver_calculate tool (via MCP/SDK) with type, ksp, stoich, and other parameters. Fallback: run equilibrium_solver.py directly.
bash# Simple Ksp: MaXb(s) <-> aM + bX python3 skills/tooluniverse-inorganic-physical-chemistry/scripts/equilibrium_solver.py \ --type ksp_simple --ksp 5.3e-27 --stoich 1:3 # Ksp + complex formation (e.g., Al(OH)3 in water with Al(OH)4- complex) python3 skills/tooluniverse-inorganic-physical-chemistry/scripts/equilibrium_solver.py \ --type ksp_kf --ksp 5.3e-27 --kf 1.1e33 --stoich 1:3 # Common ion effect (e.g., AgCl in 0.1M NaCl) python3 skills/tooluniverse-inorganic-physical-chemistry/scripts/equilibrium_solver.py \ --type common_ion --ksp 1.77e-10 --stoich 1:1 --common-ion 0.1
Key points:
ksp_kf mode solves the full charge-balance system numerically (Newton's method) — accounts for free cation, complex anion, and OH-/H+ simultaneouslyMX_b + X- <-> MX_(b+1)-, K_overall = Ksp Kfcommon_ion mode uses bisection to solve the exact Ksp expression with extra ion concentration--stoich a:b matching the salt formula (e.g., 1:3 for Al(OH)3, 1:2 for CaF2, 1:1 for AgCl)python3 skills/tooluniverse-organic-chemistry/scripts/chemistry_facts.py for reference data.| Tool | Use For | |------|---------| | PubChem_get_CID_by_compound_name | Get compound CID from name | | PubChem_get_compound_properties_by_CID | Detailed compound data by CID | | ChEMBL_search_molecules | Bioactive compounds | | PubMed_search_articles | Literature on synthesis conditions, properties | | CrystalStructure_validate tool (or crystal_validator.py fallback) | Verify crystal structure data consistency | | EquilibriumSolver_calculate tool (or equilibrium_solver.py fallback) | Ksp, complex formation, common-ion solubility |
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 6,921 | 6,936 | +0% | 1 | 1 | 0% | 1,583 | 3,307 | +109% | 0 | 0 | — |
case-02 | fail→fail | 6,601 | 6,779 | +3% | 1 | 1 | 0% | 1,517 | 3,204 | +111% | 0 | 0 | — |
case-03 | fail→fail | 7,346 | 8,356 | +14% | 1 | 1 | 0% | 1,662 | 3,037 | +83% | 0 | 0 | — |
case-04 | fail→fail | 12,795 | 13,792 | +8% | 1 | 1 | 0% | 2,357 | 3,785 | +61% | 0 | 0 | — |
case-05 | fail→fail | 15,187 | 4,138 | -73% | 1 | 1 | 0% | 1,145 | 2,528 | +121% | 0 | 0 | — |
case-06 | fail→pass | 9,148 | 16,818 | +84% | 1 | 1 | 0% | 1,879 | 4,176 | +122% | 0 | 0 | — |
case-07 | fail→fail | 7,681 | 4,285 | -44% | 1 | 1 | 0% | 1,541 | 2,395 | +55% | 0 | 0 | — |
case-08 | fail→fail | 11,456 | 12,604 | +10% | 1 | 1 | 0% | 2,414 | 4,380 | +81% | 0 | 0 | — |
case-09 | fail→fail | 10,016 | 6,023 | -40% | 1 | 1 | 0% | 1,414 | 2,875 | +103% | 0 | 0 | — |
case-10 | fail→fail | 10,262 | 11,222 | +9% | 1 | 1 | 0% | 2,009 | 3,987 | +98% | 0 | 0 | — |
case-11 | fail→fail | 3,882 | 4,141 | +7% | 1 | 1 | 0% | 1,052 | 2,562 | +144% | 0 | 0 | — |
case-12 | fail→fail | 7,945 | 5,924 | -25% | 1 | 1 | 0% | 1,844 | 3,015 | +64% | 0 | 0 | — |
case-13 | fail→pass | 9,404 | 1,673 | -82% | 1 | 1 | 0% | 641 | 1,937 | +202% | 0 | 0 | — |
case-14 | fail→fail | 15,728 | 10,854 | -31% | 1 | 1 | 0% | 2,442 | 3,790 | +55% | 0 | 0 | — |
case-15 | fail→fail | 14,423 | 13,866 | -4% | 1 | 1 | 0% | 2,705 | 4,225 | +56% | 0 | 0 | — |
case-16 | pass→pass | 11,960 | 3,626 | -70% | 1 | 1 | 0% | 2,067 | 2,243 | +9% | 0 | 0 | — |
case-17 | fail→fail | 2,770 | 2,777 | +0% | 1 | 1 | 0% | 452 | 2,139 | +373% | 0 | 0 | — |
case-18 | fail→fail | 4,433 | 3,648 | -18% | 1 | 1 | 0% | 825 | 2,461 | +198% | 0 | 0 | — |
case-19 | fail→fail | 5,773 | 3,149 | -45% | 1 | 1 | 0% | 1,155 | 2,232 | +93% | 0 | 0 | — |
case-20 | fail→fail | 13,610 | 14,120 | +4% | 1 | 1 | 0% | 2,817 | 4,512 | +60% | 0 | 0 | — |
case-21 | fail→fail | 9,438 | 10,459 | +11% | 1 | 1 | 0% | 1,829 | 3,588 | +96% | 0 | 0 | — |
case-22 | fail→fail | 8,942 | 13,275 | +48% | 1 | 1 | 0% | 1,797 | 4,476 | +149% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted, and 21 counted toward the lift figure. The other 1 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of 0 percentage points is the difference between those two pass rates over the 21 comparable cases. 1 case got worse with the skill loaded, and it is included in that figure.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.