Ranges, not promises
Each task draws from its min, likely, and max values. The likely column is one point on that distribution, not a confidence level.
Rack arrival to training
Commissioning schedule risk
A likely finish is not a deadline confidence. Propagate task uncertainty across the dependency graph, then see which work actually controls readiness.
Source boundary: the supplied post's 19-day statement is used only as a scenario target. This tool does not verify its historical, company, model, infrastructure, scale, ownership, or future claims.
Commissioning tasks
10,000
20260712
Bundled sample ready: seven tasks from rack receipt through training smoke test.
Deadline proof
Run the schedule before trusting the target.
The graph, 10,000 duration trials, and controlling-path counts are computed locally from the task table.
P50 finish
P80 finish
P95 finish
Likely plan
Target margin
Graph proof
Likely controlling path
Finish-day distribution
10,000 seeded trials
Canonical result
Read the risk
Joins wait for the slow branch
Graphlib validates the DAG. Every task starts only after all predecessors finish, so parallel work can still create a controlling join.
Criticality is observed
A task's percentage is how often it lands on the longest sampled path. That is mitigation evidence, not an arbitrary priority score.
Analyze a valid schedule to unlock its evidence export.