Function run_variant_tasks 

pub fn run_variant_tasks<F>(
    variant: &str,
    tasks: &[VariantTask],
    cmd_for: F,
    min_files: f64,
) -> Result<AgentBenchSummary, String>
where
    F: FnMut(&VariantTask, &Path) -> Result<String, String>,

The shared variant runner: for each task, copy the fixture repo, index it, ask cmd_for for the per-task shell command (this is where a variant generates its context artifact into the freshly indexed repo and returns the agent command that consumes it), then run the benchagent protocol. Aggregates the same metrics as bench agent plus first-plan accuracy and graph-query counts, and records variant.