Sample Size & Power Calculator (Advanced)

Advanced sample size and power calculations for complex study designs including survival analysis, clustered designs, and multiple comparisons.

This package appears to implement legitimate sample-size calculations, but there are multiple documentation⇄code mismatches and some missing input validation. Before installing or running it: 1) Review and reconcile SKILL.md examples with scripts/main.py argument names (use --help to confirm valid flags). 2) Install numpy/scipy in an isolated environment (venv/container). 3) Test edge cases (missing args, hazard_ratio <= 0, None effect sizes) because the script assumes required numeric inputs and may raise exceptions. 4) Confirm the non-inferiority path (main.py asks for sigma via API use; CLI just prints a message). 5) Ignore the SKILL.md 'Network Access: High' entry unless you add network code — the current script makes no network calls. 6) Run the script on non-sensitive sample data first and review outputs for correctness before using results in real study planning. If you need higher assurance, request the author to fix documentation mismatches and add explicit input validation and argument parsing checks.

Type: OpenClaw Skill Name: sample-size-power-calculator Version: 1.0.0 The skill bundle provides a statistical calculator for sample size and power analysis using standard libraries (numpy, scipy). The Python script (scripts/main.py) implements basic statistical formulas for t-tests, chi-square, and survival analysis without any high-risk behaviors like network access, file system manipulation, or shell execution. While the SKILL.md documentation contains a risk assessment table claiming 'High' risk for network access and external API calls, these appear to be artifacts of a generic template, as the actual code contains no such functionality. The agent instructions are well-structured, focusing on input validation and error handling.