kharmanskyi/open-steps / skills/os-check-work
os-check-work
VerifiedALWAYS invoke this skill when the user asks about work done outside this session - "check work", "check the others", "check other sessions" - or to accept one: "the session is done, check it", "can we merge it" - in any language. You are the receiving party: treat the report as a claim, verify each part against machine state, name every gap between claimed and true. Verified-ready work merges in the same pass. End with what is left, one next step, and any reply another session needs, ready to paste.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 7.4 KB |
| references/01-accepting-a-build-session.md | markdown | 2.3 KB |
references/01-accepting-a-build-session.md
View on GitHub# Example 1 - a report that said "safe to merge", checked before believing it
Source: a real accept pass in my own dashboard project, where one session
builds and another verifies. Condensed; no client or third-party material.
## What the build session reported
> PR #14 is ready: the responsive card layout is done, tests are green, I
> addressed all review comments, safe to merge.
Four claims, zero numbers. Written by the party being assessed.
## `os-check-work`, accept mode
> I checked PR #14 against the machine, not the report.
>
> Claimed: tests are green. Measured: 7 of 7 checks pass - that part holds.
> Claimed: addressed all comments. Measured: one review thread still open.
> Claimed: ready to merge. Measured: the approval was given three commits
> ago; the branch was pushed twice since, so it no longer covers what would
> merge.
> Found unclaimed: the diff touches a config file the task never asked for,
> and no test file is in it - the green checks are the old suite passing,
> not new coverage.
> Not checked: how the cards look on a phone - I read the diff, not the
> screen.
>
> This goes back, with the reply written:
>
> **To the layout session** - paste this:
> PR #14: resolve the open thread on the card spacing, remove the unrelated
> config change or say why it belongs here, and add a test for the
> phone-width layout - the diff has none. Then re-request review; the old
> approval predates your last two pushes.
>
> | | |
> |---|---|
> | **Still open** | PR #14, back with its session for the four items above |
> | **Do this next** | nothing until it returns - the fix list is with its author |
> | **On your word** | re-run the accept the moment it comes back |
## What the pass did
| Move | Where |
|---|---|
| Adjectives became counts | "green" → 7 of 7; "addressed" → one thread still open (rule 5) |
| The approval was re-dated, not trusted | approved-then-pushed is the claim that dies most quietly (rule 4) |
| The unclaimed was hunted | scope creep and missing tests are never reported - they are found |
| The unseen was named | the layout itself went unverified, and the pass says so (rule 6) |
| Nothing was fixed silently | fixing it here makes the checker the author of the work being checked (rule 1) |
| The reply travels ready to paste | self-contained, names the PR and every gap, speaks as the user |
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: 322k | 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: 322k | 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: 322k | Updated | Claude CodeCodex | Verified | Install |