Conciseness

Lean and dense — tables of operational guidance, no padding, no explaining of basic concepts Claude already knows (it assumes PK literacy and gives only the actionable definitions, units gotchas, and interpretation notes).

Actionability

Fully executable copy-paste examples: a complete `tu run NCA_compute_parameters` JSON payload with units, a bioavailability call with concrete numbers, and a real referenced script (`scripts/nca_from_csv.py`) with usage — covering the common cases.

Workflow Clarity

Clear 5-step sequence (prepare → run NCA → interpret → bioavailability → quality gotchas) with explicit validation thresholds (extrapolation >20%, r_squared < 0.9, F > 1) and feedback loops (>20% → report AUC0-last; bad fit → half-life unreliable; F>1 → recheck units).

Progressive Disclosure

Well-organized overview with a single clearly signaled, one-level-deep reference (the CSV script) that offloads the BLQ-handling variant rather than inlining it; clean section navigation and a related-skills pointer.

Specificity

Lists multiple concrete actions with comprehensive coverage — NCA for Cmax, Tmax, AUC (0-t and 0-∞), terminal half-life, clearance, Vd, MRT, bioavailability, plus one-compartment fitting.

Completeness

Explicitly answers both what (PK analysis / NCA for the listed 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').

Trigger Term Quality

Comprehensive natural domain terms including synonyms: 'pharmacokinetic/PK', 'concentration-time data', 'plasma/serum drug concentrations over time', 'bioavailability', 'IV + oral AUCs', 'half-life', 'clearance' — exactly what a user with PK data would say.

Distinctiveness Conflict Risk

Clear PK-NCA niche with an explicit disambiguation boundary — 'NOT for ADMET property prediction from structure (use tooluniverse-admet-prediction)' — minimizing trigger overlap with the sibling skill.