Atlassian Forge Gadget

Task Flow Simulator for Jira

Run What‑If scenarios, forecast with Monte Carlo, and explore flow efficiency — right on your Jira dashboards.

P50 / P85 / P95 Forecast Scenario A/B Control Charts Capacity Planner SLA & Epic Forecast
Features Configure Modules How it works FAQ Support

Why teams use Task Flow Simulator

What‑If & Monte Carlo

Adjust capacity %, defect ×, inflow/week. Get completion dates with P50/P85/P95, histogram and CDF.

Interactive sliders Burn‑up bands

Data Quality Index

Checks weeks(n), min/median/mean/max, CV. Warns when history is too short or unstable.

Auto‑validation Reliability hints

JQL‑aware & Safe

Source by JQL / Project / Epic keys. Time‑bounded queries to avoid unbounded scans.

/rest/api/3/search/jql Pagination

Configure (Edit screen)

Source

  • Mode: JQL / Project / Epic keys
  • History window: weeks for learning throughput (e.g., 8–26w)
  • Scope policy: All open / Updated in last N weeks
  • Include zero weeks: count silent weeks for realism

Simulation

  • Method: Throughput Monte Carlo (Cycle‑time model coming)
  • Runs: 500–10,000
  • Defaults: Capacity %, Defect ×, Inflow/week
Tip: Add multiple gadget instances (e.g., Bugs vs. Features) and compare outcomes.

Analysis modules

Monte Carlo — P50/P85/P95, histogram, CDF, burn‑up.
Control Charts & Anomalies — scatter with rolling mean and 3σ bands.
Capacity Planner — staffing hints vs. target date.
SLA / Service Level — cycle/lead percentiles vs. targets.
Epic Forecast — per‑epic P50/P85/P95.
Scope Inflow Sensitivity — effect of extra inflow/week.
“Can we finish by D?” — probability to hit a deadline.
Scenario A/B — compare several assumption sets.
CFD & Arrival/Departure — cumulative flow trends.
Risk Profile — tail risk (P95–P50), stability.
Queue Models — G/G/1, Erlang‑C (advanced).

How it works

History & Scope

History JQL collects issues with resolutiondate in the window (weekly throughput). Scope JQL counts open items to finish.

History example:
status=done AND resolutiondate >= startOfDay(-13w) AND issuetype=Bug

Simulation

Monte Carlo draws weekly throughput samples and applies capacity %, defect ×, inflow. Output → weeks‑to‑done and P‑dates.

Outputs:
P50 / P85 / P95 dates
Histogram & CDF
Burn‑up confidence bands

FAQ

Why is histogram diagonal?

If weekly throughput has a narrow distribution (e.g., ~7 every week), samples cluster tightly, so bars step diagonally.

How reliable are forecasts?

Check the Data Quality Index. More weeks and lower CV → more reliable. Use Control Charts to spot shifts.

Support

Contact your app admin or the development team. Please include:

  • Screenshot + Jira URL (redact sensitive info if needed)
  • History/Scope JQL and module used
  • Browser & OS; dev tunnel or production
© 2025 Task Flow Simulator for Jira — Atlassian Forge