One number for sprint risk.Four signals.Zero guesswork.
A 0–100 risk score combining blocked tickets, aging tasks, unassigned work and high-priority-unstarted. When a ticket goes red, the webhook fires.
The problem
You find out a sprint is in trouble on day 9 of 10.
- A ticket got blocked on day 2; nobody saw the status change.
- A ticket was unstarted high-priority on day 4; the dashboard didn't surface it.
- A ticket has been sitting in 'in progress' for 18 days, dragged forward sprint to sprint. Nobody flagged it stale.
How it works
Step by step.Nothing behind a curtain.
Task-set risk score (0–100)
Up to 30 points for the blocked-task ratio. Up to 25 for high-priority tickets still unstarted. Up to 25 for aging tickets (>14 days in current state). Up to 20 for unassigned tickets.
Board health score (0–100)
Deducts for overdue, >30% unassigned, stale (>14 days), overloaded assignees (≥6 tasks). Adds for >70% completion and active progress. Bucketed: healthy (≥75), warning (≥50), critical (<50).
Velocity volatility
Coefficient of variation across the team's last 10 sprints. Under 0.3 = 'consistent.' Above that = a team whose capacity estimate is not trustworthy.
What you get
Specific.Not marketing-department specific.
- A 0–100 risk score per task set, updated whenever a ticket moves.
- A 0–100 board health score with healthy/warning/critical bucketing.
- Stale-ticket flagging at 14 days in current state.
- Bottleneck flagging at 6+ assigned tickets per person on a single board.
- Outgoing health_alert webhook on status → blocked or priority → critical/highest. HMAC-SHA256 signed. 3 retries with exponential backoff. Auto-disable after 10 consecutive failures.
- Task Health Snapshots - a rolling 7-day log of per-ticket health (missing description, missing assignee, stuck on high priority, etc.).
Built right
One score.Four signals.Full provenance.
The sprint health score reads as one number, but every contributing signal is visible underneath. Computed from derived metrics - defensible in a security review. Threshold tuning, webhook signing and delivery behavior are all things we'd cover on a call.
Get the architecture deep-dive on a callWorks with
Know the sprint is in trouble on day 2, not day 9.
Private by default. Specific by design. Free for 2 users, forever.