Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →brycewang-stanford/drug-development-guide
End-to-end drug development pipeline from target identification to regulatory...
Passed 18 of 23 test cases with the skill, and 15 of 23 without it.
and 1 more in the full report
4 fixed14 passed either way4 still failing1 broke
Tokens, turns and time describe resource use, not a dollar price.
One of 4 test cases the skill fixed. Same model, same prompt, with and without the skill.
I am working on screening a library of small molecules for oral drug candidates. Could you provide a Python function that evaluates molecular properties—specifically weight, lipophilicity, hydrogen bond donors, and acceptors—against standard druglikeness criteria? The output should be a Python dictionary containing the total number of rule violations, an overall pass/fail boolean, detailed notes on any breached limits, and a summary assessment of oral bioavailability likelihood.
| Checked | Without skill | With skill |
|---|---|---|
| The function evaluates a molecule as passing when it has at most one rule violation. | Passed | Passed |
| The returned dictionary contains the key 'assessment'. | Missed | Passed |
The instructions your model reads when the skill loads.
A comprehensive skill covering the drug development pipeline from target identification through regulatory approval. Designed for pharmaceutical researchers, medicinal chemists, and clinical scientists conducting academic or industry research.
Target ID -> Hit Finding -> Lead Optimization -> Preclinical -> Phase I -> Phase II -> Phase III -> Regulatory Filing
(1-2 yr) (1-2 yr) (1-3 yr) (1-2 yr) (1 yr) (2 yr) (3 yr) (1-2 yr)
Total timeline: ~10-15 years | Success rate: ~5-10% from Phase I to approval
Estimated cost: $1.3B-$2.8B per approved drug (DiMasi et al., 2016)
import pandas as pd
from scipy import stats
def differential_expression_analysis(expression_data: pd.DataFrame,
disease_group: list[str],
control_group: list[str],
fdr_threshold: float = 0.05) -> pd.DataFrame:
"""
Identify differentially expressed genes as potential drug targets.
Args:
expression_data: Gene x Sample expression matrix
disease_group: Sample IDs in disease condition
control_group: Sample IDs in control condition
fdr_threshold: False discovery rate threshold
"""
results = []
for gene in expression_data.index:
disease_vals = expression_data.loc[gene, disease_group]
control_vals = expression_data.loc[gene, control_group]
t_stat, p_value = stats.ttest_ind(disease_vals, control_vals)
test cases passed, vs 15 / 23 without
More skills from the same source and nearby in the registry, with their headline benchmark lift.