- name
SKILL.md
Файл навыка остаётся в репозитории издателя.
Вставьте скопированную команду в терминал. Она установит этот SKILL.md в выбранного агента.
npx skills add kharmanskyi/open-steps --skill os-whats-next -a cursor -yВставьте скопированную команду в терминал. Она установит этот SKILL.md в выбранного агента.
npx skills add kharmanskyi/open-steps --skill os-whats-next -a claude-code -yВставьте скопированную команду в терминал. Она установит этот SKILL.md в выбранного агента.
npx skills add kharmanskyi/open-steps --skill os-whats-next -a codex -yВставьте скопированную команду в терминал. Она установит этот SKILL.md в выбранного агента.
npx skills add kharmanskyi/open-steps --skill os-whats-next -a grok -yОтправьте скопированное сообщение в чат Grok Bot. Бот загрузит этот SKILL.md с GitHub и добавит его к навыкам, общим для всех ваших ботов.
Install this skill: https://github.com/kharmanskyi/open-steps/tree/HEAD/skills/os-whats-nextОт издателя
Answer "what do we do now" with the next move, not a map. The user does not
need the dependency graph - they need what got finished, what to take next,
and why, in plain words. This skill decides; os-step-by-step walks the user
through their part; os-done-or-not reports what came of it.
Write in the language the user speaks in this session, detected from the conversation. Commands, file names and identifiers stay English.
Triggers: the description above, plus a session just ended wanting a next move.
Stop as soon as you can answer.
The last report - ~/.claude/open-steps/reports/<project>/latest.md:
a handover written for exactly this moment.
Local state - uncommitted changes, unpushed commits, current branch.
Open pull requests - one call:
gh pr list --json number,title,mergeStateStatus,reviewDecision,isDraft.
The backlog - always. The issue tracker when one is already connected
(never authenticate or install one), otherwise task files in the repo:
BIG-PICTURE.md, PLAN.md, TODO.md, docs/plan*. Next work comes from
the backlog, not from imagination. No backlog anywhere → say so.
Where os-big-picture keeps that file, only the "What is next" section
is the backlog. The feature table above it is an inventory, and a row
that has not changed in months is a finished feature, not a task. Reading
work out of it is inventing work, which rule 4 below forbids. A row under
"Worth retiring" is a real candidate, but it is a decision to put to the
user, never a task to start.
That file carries its own age: a measured date at the top and a date beside
every stage. Never take a number out of it. Take only the "What is next"
list, the "Worth retiring" rows, and each Stage with the date beside it.
Before you put a "Worth retiring" row to the user, measure it again. From the project root, run
bash <this skill's folder>/../os-big-picture/scripts/census.sh .. The .
is the project, not the skill folder. If the script is missing or does not
run, say so and call that row "not checked". When the stage dates are
months behind the newest commit, say how old they are. Dates that stopped
moving mean the reports stopped, not that the work did.
Say which sources you did not read: an unread source is not an empty source.
A pull request with green checks and an approval is not a decision - it is
unfinished business. Verify it through os-check-work's accept rules and
merge it in this same pass. Two things stop the merge: a failed claim, and a
task instruction that merges happen on command only - an orchestrator may own
the merge. Report it as done, never as a question.
Blocked work gets no section of its own. Fold it into the reasoning, in plain words - "X waits on an outside check; I watch it" - the user trusts the recommendation, not the graph.
<Lead: one sentence on where things stand - including what this pass merged.>
**I can do alone:** <up to three items, five words of why each>
**Needs you:** <up to three items, one line each - or drop the list>
**Next I take: <the one task> - <plain words: what it closes or unblocks>.**
<One line: what was not checked.>
When a quick small win and a big item are both real candidates, offer the choice through your tool's question picker where it has one (in Claude Code, AskUserQuestion). Give two to four options, the recommended one first and marked. Where there is no picker, write the same question and options as plain text, the recommended one first and marked. On the pick, prepare the launch: a prompt complete enough to paste or a command complete enough to run, and one line saying what comes out. Never run it yourself.
Before offering to start several, prove they will not collide - all three:
Any check failing → one at a time, saying which failed. All passing → say so. Never claim parallel safety you did not verify - "I did not check" is honest; a collision discovered mid-run is not. Where the project isolates parallel work - a working copy per task, separate container projects or ports - name that as the precondition instead of assuming it.
os-step-by-step, never explained inline.>-
---name: os-whats-nextdescription: >----# skills/os-whats-next/SKILL.mdFAQ
Скачайте skills/os-whats-next/SKILL.md у издателя или скопируйте путь и откройте репозиторий. PlugMCP только показывает навык в каталоге и не хранит файл.
Издатель хранит его в https://github.com/kharmanskyi/open-steps.
Claude, Cursor, Codex и Grok Build могут читать SKILL.md, который вы добавляете из репозитория издателя. Grok Bot устанавливает его, когда вы отправляете в чат ссылку на папку навыка на GitHub.