@plan

Project management library for auditable notebooks. CPM scheduling, PERT estimation, Monte Carlo simulation, earned value management, and analysis functions.

usage

const plan = await load("@plan");

modules

| File | Description | |------|-------------| | calendar.js | Working day calculations, holidays, blocked periods | | pert.js | PERT three-point estimation | | graph.js | Topological sort, cycle detection, predecessors/successors | | schedule.js | CPM forward/backward pass, critical path, float | | resource.js | Resource conflict detection, leveling | | scurve.js | S-curve generation (planned/actual/forecast) | | evm.js | Earned value management (BAC, EV, PV, AC, SPI, CPI) | | analysis.js | 16 analysis functions (see below) | | workflow.js | Workflow templates, stage gates, throughput | | montecarlo.js | Monte Carlo simulation with sensitivity tracking | | render.js | SVG renderers (Gantt, S-curve, histogram, tornado, etc.) | | xlsx.js | Excel export | | holidays-br.js | Brazilian holidays (federal, state, municipal) |

analysis functions

• Schedule: whatIf, delayImpact, nearCritical, slackBudget, scopeDrift, bufferStatus
• Resource: busFactor, switchingOverhead, meetingCost, constraint
• Math models: brooksLaw, littlesLaw, multiProjectFragmentation
• Progress: burndown, health, compress

renderers

gantt, scurvePlot, resourceHistogram, stageGateView, workflowDiagram, monteCarloPlot, deadlineRiskPlot, tornadoPlot, burndownPlot

brazilian holidays

// Federal + state + municipal
const cal = plan.brazilCalendar(2026, 2027, { municipality: "belo horizonte-MG" });
cal.blocked = [{ start: "2026-04-06", end: "2026-04-17", label: "Vacation" }];

// Options: carnival, corpusChristi, optional (pontos facultativos)
plan.brazilMunicipalities(); // list available municipalities

Covers all 27 state capitals + Parauapebas, Canaa dos Carajas, Maraba (PA).

build

node ext/plan/build.js

Concatenates src/*.js into index.js, stripping ES module syntax.

roadmap

rework duration scaling

Current rework transitions in Monte Carlo re-run tasks at full PERT duration. In practice, rework is faster than the original pass (data exists, parameters are tuned, downstream artifacts just need updating). Enhancement:

reworkTransitions: [{
  from: "validation", to: "estimation",
  probability: 0.3,
  durationScale: { estimation: 0.4, validation: 0.5 }
}]

durationScale would multiply the sampled PERT duration on rework iterations, per task. Tasks not listed default to 1.0 (or a configurable defaultReworkScale). This models the reality that re-estimation after a validation failure takes ~40% of original time, not 100%.

not yet implemented (need historical data)

darkTime, gateMetrics, learningCurve, reworkAmplification, referenceClass, overheadRatio, outsourcingAnalysis, fragmentation, utilizationCurve — all require completedInstances / stageHistory data.