🔒Connected to local proxy — your PAT stays on the server.
⚠️Forecast assumptions:
Items are sized like past work · Team delivers at its historical rate ·
WIP stays disciplined (don't start all tickets at once) ·
History window is representative of the future.
Cycle-time start uses ActivatedDate (approximate where missing).
⚙️ Forecast Settings
▼ Show
States that count as "completed" — varies by process template
Forecast grain — mix Epics/Features in with caution
How far back to pull completed items for throughput
Bucket size for the throughput chart and simulation
More trials = smoother percentile curves, slower browser
IQR fence multiplier for outlier detection
Count of tickets in the batch
"How many by this date?" mode
⚠️
Completed (window)
—
items
Avg Throughput
—
per week
Max in one interval
—
best week
Min in one interval
—
slowest week
Buckets sampled
—
intervals with data
📊 Historical Throughput
🔍 Outlier Insights
ⓘ Detection uses Tukey IQR fence (or MAD for small samples). Heuristic labels are likely explanations, not certainties.
This tool cannot see team holidays or PTO — a quiet holiday week looks identical to a capacity problem.
Week Detail
🗓️ How Long to Complete 20 Items?
Commit externally to the 85% date
⏳
Run the forecast to see completion-date estimates.
📉 Completion-Date Distribution (CDF) — % of simulations done by date
🎯 How Many Items by —?
Rank tickets by urgency and commit to the top X
📊 Simulated Completion Count Distribution — % of simulations completing exactly N items