Fill all four to see the focused-hours capacity.
Velocity averager (points from recent sprints)
List the points completed in your recent sprints to see the average.
Nothing to paste? — a payments squad's Sprint 25: eleven stories with a 13-pt splitting candidate, a story the team refused to size, a dependency parked with another team, a contractual penalty-clause date, a carry-over, PTO and a sales promise squeezed on top. Or , rendered from a saved reply with no run and no charge.
Paste the backlog
Story lines plus the scenario: team, sprint dates, velocity history, PTO, carry-over, pressure. A CSV or TSV straight out of Jira, Linear or a spreadsheet works too — it is converted into story lines before scanning. The story scanner sizes the ask up free in your browser — point totals, unestimated and off-scale flags, a velocity read. No sign-in needed for that.
Say what you need
Full sprint plan, estimation review, capacity check, goal setting, or scope negotiation — plus any notes. A commitment the numbers say cannot fit is reported whatever the task; the task sets where the depth goes.
Walk into planning
The capacity named with its velocity arithmetic, a goal the stories add up to, a backlog where every story earns its slot, the cut list with what unblocks each story — and the risks, actions and open questions. Your browser then totals the committed points, checks them against the pasted velocity, and names any pasted story that landed in neither table.
Take it where it is going
Copy the goal for the board field, download the issue CSV that imports into Jira or Linear, copy the actions and open questions as a planning agenda, and copy the cut list back out as story lines to start the next sprint from. Every plan is saved to your account, so the next run can be diffed against this one.
Does Sprint Desk cost anything?
The story scanner, the tracker-export conversion, the capacity calculator, the velocity averager, the example plan and every export are free and need no account — they run entirely in your browser. Only the model-written plan is metered, and the page shows what a run reserves before you start it and what it actually charged afterwards. If your balance is below what the model needs, the run button is disabled with the shortfall named rather than failing after you press it.
Can I paste a Jira or Linear export instead of typing story lines?
Yes. Paste the CSV or TSV as-is. If it carries a header row naming an issue key and a summary column — plus story points, assignee or a dependency column when you have them — those rows are rewritten into the story-line shape before anything is scanned or sent. The scan panel says how many rows were converted. Your text box keeps your original text; only what is sent is rewritten.
How do I know the plan's numbers are real?
Two ways. The prompt requires every number to be traceable to your paste, with the arithmetic stated in the same bullet and no industry defaults quietly filled in. Then your browser checks the result independently: it totals the points in the committed and deferred tables, compares the committed total against the velocity history in your paste, and flags a plan tagged committable whose own rows do not fit. It also names any story you pasted that appears in neither table, and any row naming a story your paste never contained.
What happens when I run it again after changing something?
Each plan is saved to your SkillSafe account through the platform's per-user store, with this browser as a mirror, so your plans follow you to another machine. Pick an earlier plan as a baseline and the page diffs the two on ticket ids and points: what moved into the commitment, what moved out to the cut list, what is new, what is in neither table any more, and how the commitment tag, the confidence and the committed total changed.
What leaves my browser?
Nothing at all until you press Plan the sprint. The scan, the conversion and both calculators are local. When you do run, the paste, your notes, the chosen task and the browser scan's summary are sent to the model through SkillSafe. If the paste is very long the middle is dropped rather than the end — the ranked stories at the top and the velocity, PTO and pressure notes at the bottom both survive — and the cut is declared in the request so the plan says it saw a partial paste.
Which estimation scale does it assume?
The Modified Fibonacci scale — 1, 2, 3, 5, 8, 13, 20. A story pasted with an off-scale value keeps the value you gave it and gets flagged rather than silently rounded. Anything at 13 or above is treated as a splitting candidate, and the plan names the seam where the description supports one. Capacity is anchored on your velocity history when you paste one, or on team, days, focused hours and a stated focus factor between 0.6 and 0.8 when you do not.