Conciseness

The body is dense and table-driven with essentially no padding, but the interpretation table re-states basics Claude already knows ("Cmax/Tmax — Peak concentration & time to peak", "MRT — Mean residence time") and includes one flourish ("NCA is the model-independent workhorse"). That places it at the efficient-with-minor-trims level 4 rather than the every-token-earns-its-place level 5.

Actionability

Commands are copy-paste executable with complete JSON payloads ("tu run NCA_compute_parameters '{...}'", "tu run NCA_calculate_bioavailability '{...}'"), a real local script (scripts/nca_from_csv.py) is given for CSV input, and the bioavailability formula is explicit. This matches the fully-executable level-5 anchor; level 4 would require gaps in the covered cases.

Workflow Clarity

Steps 1–5 are clearly sequenced (prepare data → run NCA → interpret → bioavailability → quality gotchas) with explicit validation checkpoints and corrective actions: ">20% → AUC0-∞ ... is unreliable; report AUC0-last instead", "Trust only if r_squared_terminal_fit ≥ ~0.95", "F > 1 signals a data/dosing error (recheck units and doses)". This matches the explicit-validation-with-error-recovery level-5 anchor; the operations are analytical, so no destructive-batch cap applies.

Progressive Disclosure

The single bundle file (scripts/nca_from_csv.py) is real, referenced clearly in the body, and one level deep; the body is well-sectioned with nothing inlined that belongs in a separate file. The interpretation table is compact enough to stay inline, and the related-skills section aids navigation, fitting the level-5 anchor (level 4 would need misplaced content or unclearly signaled references, which are absent).

Specificity

The description enumerates concrete analyses — "non-compartmental analysis (NCA) for Cmax, Tmax, AUC (0-t and 0-∞), terminal half-life, clearance (CL), volume of distribution (Vd), MRT, and absolute bioavailability (F). Also one-compartment fitting" — matching the comprehensive multi-action anchor rather than the 'minor gaps' level 4.

Completeness

It explicitly answers what (NCA for the named parameters, one-compartment fitting) and when ("Use when you have plasma/serum drug concentrations over time after a dose and need PK parameters, or to compute bioavailability from IV + oral AUCs"), with concrete trigger phrases exactly like the level-5 anchor.

Trigger Term Quality

It covers natural user phrasings and their synonyms: "pharmacokinetic (PK)", "plasma/serum drug concentrations over time", "PK parameters", "bioavailability", "half-life", "clearance", "Cmax", "AUC", "IV + oral". Level 4 would require missing natural terms, but no common variant is absent.

Distinctiveness Conflict Risk

The domain is highly specific and it contains an explicit negative boundary — "NOT for ADMET property prediction from structure (use tooluniverse-admet-prediction)" — which removes the main overlap risk, fitting the clear-niche level-5 anchor.