kharmanskyi/open-steps / skills/os-whats-next
os-whats-next
VerifiedALWAYS invoke this skill when the user asks what to do next, what is left, or what is blocked - "what's next", "what now", "what should we work on", "anything I can do" - in any language. This skill picks the next piece of work; when the user asks HOW to do a thing or says they do not understand what to do, that is os-step-by-step. Reads the last report, local changes, open pull requests and always the backlog. First finishes what is finished: verified-ready pull requests merge in the same pass. Then sorts the rest into doable-alone and needs-you, ending with one recommended next task in plain words - what it closes or unblocks. Never invents tasks.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 7.4 KB |
| references/01-end-of-day.md | markdown | 2.5 KB |
references/01-end-of-day.md
View on GitHub# Example 1 - "what's next?" answered with a map instead of a move
Source: a real end-of-day moment in my own exam-practice platform. Condensed;
no client or third-party material.
## What the agent answered
> Current state: feat/scoring-fix is merged, feat/results-email is WIP on my
> branch (email template done, SMTP wiring pending), PR #31 (mobile layout)
> has green CI awaiting review. Backlog: PostHog events not instrumented,
> testimonials section still stubbed, pricing page copy TODO, oral-exam
> simulator needs the TTS voice swapped. Also the Vercel preview for #31 404s
> on the results route, investigating. What should I prioritize?
Everything true, nothing usable. The person asked for the next move and got
the dependency graph, ending in the question they had just asked the agent.
## `os-whats-next`
> The scoring fix landed earlier, so practice tests grade correctly again.
> Nothing was ready to finish in this pass: the mobile layout has green checks
> but no review yet, and its preview breaks on the results page, so it stays
> open until that is traced.
>
> **I can do alone:** trace the broken results page on the mobile-layout
> preview (its pull request) - it blocks the review; finish the results email
> (my branch) - no result reaches the student today; wire the two analytics
> events (backlog) - the launch decision needs them.
> **Needs you:** two real student testimonials (backlog) - only you can ask
> them; the ads budget stays parked until testimonials and analytics exist -
> your earlier rule.
>
> **Next I take: the broken results page on the mobile-layout preview - small,
> and until it is fixed nobody can review that work.**
> I did not check the ad account or the payment provider's dashboard.
## What the rewrite did
| Move | Where |
|---|---|
| Nothing was merged that had not earned it | green checks with no review and a broken preview is a claim, not a finish - it stays open with its blocker named (step 2, last gotcha) |
| The graph became two lists | doable-alone and needs-you; the broken preview became an item with its why, not a section of its own |
| Every item names its source | the pull request, my branch, the backlog, the user's own standing rule on ads - nothing invented (rule 2) |
| One move, with its why | what it unblocks, in the product's words, not the branch's |
| The unread was named | an unchecked source is not an empty source (rule 5) |
| The closing question disappeared | the skill decides; asking back "what should I prioritize" is the failure being fixed |
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness. | Stars: 323k | Updated | Claude CodeCodex | Verified | Install |
| dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies | Stars: 323k | Updated | Claude CodeCodex | Verified | Install |
| executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available | Stars: 323k | Updated | Claude CodeCodex | Verified | Install |