keaven

Guide users through classical group sequential trial design using the gsDesign R package. Use this skill when the user asks about: group sequential boundaries, spending functions (sfLDOF, sfHSD, sfPoints), sample size for time-to-event or binomial trials, gsDesign objects, plotting group sequential bounds, gsSurvPower for power computation, or harm bounds (test.type 7/8).

Guide users through next-generation group sequential design using the gsDesign2 R package. Use this skill when the user asks about: gs_design_ahr, gs_power_ahr, gs_update_ahr, sequential_pval, average hazard ratio designs, non-proportional hazards, piecewise enrollment/failure rates, spending time, or information fraction computation.

Guide users through clinical trial simulation using the simtrial R package. Use this skill when the user asks about: simulating survival trials, simfix, sim_pw_surv, cutting data at calendar or event times, weighted logrank tests, MaxCombo tests, or simulation-based power.

Guide users through classical group sequential trial design using the gsDesign R package. Use this skill when the user asks about: group sequential boundaries, spending functions (sfLDOF, sfHSD, sfPoints), sample size for time-to-event or binomial trials, gsDesign objects, plotting group sequential bounds, gsSurvPower for power computation, or harm bounds (test.type 7/8).

Guide users through next-generation group sequential design using the gsDesign2 R package. Use this skill when the user asks about: gs_design_ahr, gs_power_ahr, gs_update_ahr, sequential_pval, average hazard ratio designs, non-proportional hazards, piecewise enrollment/failure rates, spending time, or information fraction computation.

Guide users through clinical trial simulation using the simtrial R package. Use this skill when the user asks about: simulating survival trials, simfix, sim_pw_surv, cutting data at calendar or event times, weighted logrank tests, MaxCombo tests, or simulation-based power.

Guide users through group sequential design with graphical multiplicity control using the graphicalMCP and gsDesign2 R packages. Use this skill whenever the user asks about: group sequential designs with multiple hypotheses, graphical multiplicity testing, sequential p-values with gsDesign2, combining graphicalMCP with gsDesign2, clinical trial designs with multiple endpoints and populations, Maurer-Bretz procedures, alpha-spending with multiplicity graphs, or adapting the gMCPLite vignette template. Also trigger when users mention spending time, information fraction, or sequential p-values in the context of group sequential or graphical testing.

Guide users through graphical MCP procedures using the gMCPLite R package (legacy). Use this skill when the user asks about: hGraph for multiplicity graph visualization, gMCP for closed testing, or legacy graphical MCP workflows. For new projects, prefer graphicalMCP.

Choose the appropriate gsDesign survival or exact-binomial workflow when requests involve calendar-time interim analyses, event- or information-driven survival looks, seasonal rare-event exact-binomial monitoring, or explicit randomization ratios; covers when to use gsSurvCalendar(), gsSurv(), simBinomialSeasonalExact(), and toBinomialExact().

Build and document exact binomial group sequential designs with spending in gsDesign. Use when the user mentions toBinomialExact, gsBinomialExact, repeatedPValueBinomialExact, sequentialPValueBinomialExact, simBinomialSeasonalExact, vaccine/prevention efficacy with seasonal looks, or blinded information-based adaptation with fixed spending fractions.

Design, simulate, and adapt clinical trials with negative binomial recurrent event endpoints using the gsDesignNB R package. Use this skill when the task involves: NB sample size or power, event gaps and Jensen correction, calendar-time group sequential design, blinded or unblinded information estimation, sample size re-estimation (SSR), recurrent-event simulation (constant or seasonal rates), interim data cuts, completer-based analyses, or non-inferiority/super-superiority designs for recurrent events.

Route gsDesignNB simulation requests to the right package functions and vignettes. Use when working on recurrent-event simulation, group sequential simulation, sample size re-estimation, blinded information, or when the user needs help choosing between nb_sim, sim_gs_nbinom, sim_ssr_nbinom, and related workflows.

Draft workflow for general vldaft R/C/Rust package development, backend parity, compiled interface changes, tests, package checks, and maintenance tasks. Use for ordinary implementation and verification work; use vldaft-model-development instead when the task starts from a new statistical model prompt.

Use when turning a statistical survival-model prompt into vldaft package code, especially new AFT distributions, cure models, censoring or truncation behavior, parameter/link definitions, likelihood/score implementations, C/Rust backend changes, gamlss2 family support, simulation checks, backend parity tests, and model-facing documentation.