agentleFS
mehboobali98mehboobali98/rmine-skills/skills/sprint-plan/SKILL.md

sprint-plan

Skill0 starsChanged 10 days agoMITView on GitHub

Check whether a sprint or monthly bucket fits the team — remaining estimated work per assignee against each person's capacity, with the tickets nobody has estimated called out rather than guessed. Use when the user asks if a sprint is overbooked, who has too much on, whether the team can finish a version or bucket, or wants capacity planning for a Redmine target version.

36 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use balance-check, team-combat and 10 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • CoralGameSkill

    balance-check

    Analyzes game balance data files, formulas, and configuration to identify outliers, broken progressions, degenerate strategies, and economy imbalances. Use after modifying any balance-related data or design.

    Also in 5 projects that use this

  • CoralGameSkill

    team-combat

    Orchestrate the combat team: coordinates game-designer, gameplay-programmer, ai-programmer, technical-artist, sound-designer, and qa-tester to design, implement, and validate a combat feature end-to-end.

    Also in 5 projects that use this

  • +10 more from CoralGame

The file

We split this skill into the rules it gives an agent

4 rules. 0 are common, 1 is distinctive, and 3 are very specific to this project.

  • 0

    Common

    Most of the 35 edited versions we read keep these.

  • 1

    Distinctive

    This author's own choices. Worth a look.

  • 3

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Only adjust for estimation drift if /calibrate's own bar is met on the current corpus: run its script and read the bar line for each mode, …

    line 94
  2. Distinctive:

    If the bar isn't met, report unadjusted numbers and say plainly why, e.g.

    line 104
  3. Very specific:

    If it is met, the adjustment is one extra line: the team's remaining work from /estimate files priced in that mode, multiplied by that …

    line 108
  4. Very specific:

    Never rewrite a per-ticket estimate with it — /calibrate forbids blanket multipliers for good reason.

    line 113

“Common” is measured: kept by at least 6 in 10 of the 35 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Sprint plan

/calibrate says its weighted ratio "is what a sprint's capacity planning cares about", and nothing did capacity planning. This skill does: it takes a Redmine target version (a sprint, or a monthly bucket), works out each assignee's remaining estimated work, and sets it against the hours they actually have in the period.

It is read-only. It never reassigns, re-versions or re-estimates a ticket. What to move is a conversation for the team; this skill supplies the numbers for it.

A team total hides the problem this exists to find. A version can sit comfortably under team capacity while several people on it are at double theirs. Always lead with the per-person view.

0. Preflight

rmine whoami

Filtering by version needs rmine v0.6.0 or later.

1. Which version, and which period

Take a version name or ID and a project. If the user names neither, list the open versions with rmine project versions <project>; for monthly buckets the current month's is the usual answer.

The period is when the work has to be done. Default to today until the end of the month (or the version's due date, if it has one), and say which dates you used. Planning the rest of a month on its last Friday is a real question, so don't silently widen it.

2. Capacity

Capacity is working days in the period, minus each person's leave, times hours per day. Ask once, in one question, for what the numbers can't know:

  • Hours per day a person spends on ticket work. The default is 8; after meetings and reviews it is often nearer 6, and that alone can flip a verdict.
  • Weekend days (default Saturday and Sunday) and public holidays in the period.
  • Leave: who is off, and for how many working days. Use the names exactly as Redmine shows them; the script lists any leave name it didn't match.
  • Anyone only partly on this version. The script assumes everyone assigned here works only on it. Express part-time as leave days (half-time for 10 days is 5 days' leave) and say you did.

3. Run it

python3 "${CLAUDE_PLUGIN_ROOT}/skills/sprint-plan/scripts/plan.py" \
  --project <project> --version "<version>" --estimates <repo>/estimates \
  --from <start> --until <end> --hours-per-day <h> \
  --leave "<Name>=<days>" --holiday <YYYY-MM-DD> --json

Pass --estimates whenever the product repo has a committed estimates/ folder. For each open ticket the script takes the /estimate file's Total if there is one, and Redmine's Estimated time otherwise, and flags the ticket when the two differ by more than 25%. It subtracts the time already logged, leaving out leave booked against the ticket (the same rule /calibrate uses), to get what remains.

4. Read it before reporting it

  • over means remaining estimated work already exceeds capacity, before counting anything unestimated. That is a firm finding.
  • unknown, N unestimated means the person has tickets with no estimate, so their real load can't be known. Never fill the gap with a guess or an average: an invented number in a capacity plan reads as a measured one. List the tickets instead; getting them estimated (with /estimate, or in Redmine) is the action.
  • tight is above 85%, with no room for the unplanned work every sprint has.
  • no one to do it is unassigned work. It needs an owner before it can be planned at all.
  • Overruns (time logged exceeds the estimate) count as 0 remaining, which is almost always wrong — the ticket is still open. Treat each one as unestimated work and say so; the estimate needs revisiting.
  • Disagreements between an /estimate file and Redmine's field mean one of them is stale. The plan used the file; name the tickets so someone fixes the other.

5. Drift

Only adjust for estimation drift if /calibrate's own bar is met on the current corpus: run its script and read the bar line for each mode, plus the tickets flagged as well below the rest. The bar is per mode, so a mode's weighted ratio applies only to estimates priced in that mode (their Mode: line), never to the others.

python3 "${CLAUDE_PLUGIN_ROOT}/skills/calibrate/scripts/actuals.py" <repo>/estimates

If the bar isn't met, report unadjusted numbers and say plainly why, e.g. "the estimate corpus has 6 tickets; /calibrate needs 20 before its drift means anything." That sentence is part of the report, not a footnote.

If it is met, the adjustment is one extra line: the team's remaining work from /estimate files priced in that mode, multiplied by that mode's weighted ratio. It never touches estimates that came from Redmine's field, because the drift was measured on /estimate output and says nothing about how Redmine's numbers were made. Never rewrite a per-ticket estimate with it — /calibrate forbids blanket multipliers for good reason.

6. Report

Lead with the people who are over, then unknown, then tight, each with their remaining hours, capacity and the tickets that make up the load. Then the team line, the unassigned tickets, overruns, disagreements, and the drift sentence from step 5.

End with what would change the picture, as suggestions for the user to act on: tickets to estimate, overruns to re-estimate, work that could move from an over person to an ok one. Name tickets, not categories.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.