operator-backlog-scrum
Skill1 starsChanged 4 months agoMITView on GitHub
Report what changed since the last check-in - commits, backlog movement, what is ready to work and what is waiting on the product owner. Use when the user asks what has been done lately, at the start of a working session, or before a handoff.
In this repository23 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- .github/copilot-instructions.mdCopilot instructions
- templates/copilot-instructions.mdCopilot instructions
- speckit-analyzeSkill
- speckit-checklistSkill
- speckit-clarifySkill
- speckit-constitutionSkill
- speckit-convergeSkill
- speckit-implementSkill
- speckit-planSkill
- speckit-specifySkill
- speckit-tasksSkill
- speckit-taskstoissuesSkill
- backlogSkill
- code-intelligenceSkill
- field-notesSkill
- operator-backlog-newitemSkill
- operator-backlog-refinementSkill
- peer-agentsSkill
- request-auditSkill
- spec-drivenSkill
- worktreesSkill
The file
We split this skill into the rules it gives an agent
3 rules. About 0 are common, 2 are distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈2
Distinctive · estimate
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 36
If the recorded commit no longer resolves — a rewritten history, a fresh clone — the report says so and covers everything rather than …
- Distinctive: line 44
Do not paste it. Read it and say, in a few lines:
- Very specific: line 63
When an item is done, close it in the same commit that does the work: set status, closed and commit, and update the linked spec.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Check in
backlog scrum # report, and advance the watermark
backlog scrum --peek # report without advancing it"What has been done lately?" is a question agents answer badly, and the reason
is structural rather than a matter of effort: closed work is answerable from
git log, but open and in-flight work were answerable from nothing, and an
agent asked to summarise them reaches for its own context — which is a
re-summary of a re-summary and cannot be checked against anything.
backlog scrum reads git and the tracked backlog instead. Everything in the
report has a file or a commit behind it.
The watermark
"Since the last check-in" needs a durable boundary. This one lives in the per-project directory outside the repository, keyed on the primary checkout, so every worktree of a project shares it and no session's death resets it. It is deliberately not session state: session state is the artifact this machine has measured not to survive.
A plain backlog scrum advances the watermark — it consumes the period, like
reading a mailbox. Use --peek when you want the report but the human has
not seen it yet, and in particular when you are drafting a handoff. Reporting a
period into a file nobody has read yet, and then marking it read, is how a week
of work goes unmentioned.
If the recorded commit no longer resolves — a rewritten history, a fresh clone — the report says so and covers everything rather than silently covering nothing. If the watermark cannot be written, the command fails loudly, because a check-in that reports and does not advance repeats itself, and a repeated report looks exactly like a quiet week.
Reading the report to the user
Do not paste it. Read it and say, in a few lines:
- What shipped — commits, in the user's terms rather than subject lines.
- What moved in the backlog — filed, approved, closed.
- What is ready to work, and what you propose to pick up next.
- What is waiting on them — items awaiting approval, and any question blocking you. This is the part they can act on; lead with it if it is not empty.
- Anything under "Caveats on this report" — those are the parts the tool could not verify. Never drop them; a report that quietly omits its own failures is the thing this command exists to replace.
Ending the session on a completed item
The other half of the user's complaint is that sessions end wherever the context ran out rather than at a shippable increment. So:
- Pick the next item from
backlog ready— not from your own sense of what is important. That is what the gate is for. - When an item is done, close it in the same commit that does the work: set
status,closedandcommit, and update the linked spec. A close landing separately from its fix is a window in which the backlog is wrong. - Then check in again, and write the handoff. The handoff says what is in flight; the backlog says what is outstanding. They are not the same document and neither substitutes for the other.
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
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.

