windows-test
Skill5 starsChanged 24 days agoApache-2.0View on GitHub
Runs HyperHand's automated tests on the Windows host from WSL — go vet and the race-enabled tests through scripts/test-windows.sh, focused runs, parallel runs from several worktrees without overwriting each other, and reproducing a defect on the unfixed code before fixing it. Use after changing Go code or tests, before committing, when a reviewer reports a defect, and in every subagent that writes code.
- Deletes or force-pushes
In this repository5 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- dev-installSkill
- releaseSkill
- vm-acceptanceSkill
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 33
Before committing, run everything once:
- Distinctive: line 46
Write the test that shows the failure and run it: it must fail for the reported reason (check the message in the log, not only the exit …
- Very specific: line 48
When the test was written after the fix, prove it still catches the defect: git stash push <fixed files>, run the test (fails), git stash …
≈ “Common” is an estimate: at least 1 in 5 of the 372,429 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
Running the automated tests
What the script does and why it runs on the host: Building. All commands run from the repository (or worktree) root in WSL. The tests never touch the installed service or a VM; acceptance on a real VM is the vm-acceptance skill.
1. Pick a test directory
The script compiles into %TEMP%\<name> on the host and deletes that directory first. Several runs at once (subagents in their own worktrees, a reviewer and an implementer) need different names, or they delete each other's executables:
export HH_TEST_NAME=hyperhand-tests-<worktree or task> # letters, digits, '.', '_' or '-'; not . or ..A single run on the main checkout may keep the default hyperhand-tests. Delete a directory you named when your work ends, in the shell where HH_TEST_NAME is set; ${HH_TEST_NAME:?} stops the command when it is not, which would otherwise delete all of %TEMP%:
rm -rf "$(wslpath "$(/mnt/c/Windows/System32/WindowsPowerShell/v1.0/powershell.exe -NoProfile -Command '[IO.Path]::GetTempPath()' | tr -d '\r')")${HH_TEST_NAME:?}"2. Run
While working on one area, run the tests that cover it; extra arguments go to every test executable:
scripts/test-windows.sh -test.run 'TestCheckUsable|TestClickAutoActivation'Before committing, run everything once:
scripts/test-windows.sh
python3 -m unittest discover -s client # when client/ changedExit code 0 means vet and every package passed. A FAIL line names the package and its log %TEMP%\<name>\<package>.log; read the --- FAIL and following lines there. gofmt -l cmd internal must print nothing.
3. Reproduce a defect before fixing it
A reported defect is fixed only after a test fails on the unfixed code (AGENTS.md):
- Write the test that shows the failure and run it: it must fail for the reported reason (check the message in the log, not only the exit code).
- Make the fix, run the test again: it passes.
- When the test was written after the fix, prove it still catches the defect:
git stash push <fixed files>, run the test (fails),git stash pop, run it again (passes). Stash only the fix, never the new test.
4. Known non-product failures
- Three agent tests that need the host desktop (
TestWindowAt,TestFocusedHelper,TestControlHintNativeIdentityAndFallback) skip while the host session is locked;-test.vlists skips. - Windows Go cannot build a
\\wsl.localhostcheckout; always go through the script, nevergo testfrom Windows directly.
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.

