Files
deusdata--codebase-memory-mcp/scripts/test-windows.ps1
Martin Vogel a54ea95719
DCO / dco (push) Has been cancelled
fix(windows): ship one binary — remove the launcher stub flagged as a dropper
Windows shipped a PAIR: a small permanent launcher (codebase-memory-mcp.exe)
plus the real product binary (codebase-memory-mcp.payload.exe). The launcher
existed for exactly one reason — a running .exe cannot replace its own image
on Windows, so an in-process self-update needs a second resident binary to do
the swap.

That stub is statically indistinguishable from a dropper: a small, unsigned,
zero-prevalence PE whose whole job is verify-and-execute another binary.
Defender's ML scored it Trojan:Win32/Wacatac.B!ml and blocked the v0.9.1-rc.1
release at the VirusTotal gate. It is not fixable in our code on x64 —
bcrypt-free, stripped, VERSIONINFO'd, minimal-resource and even
resource-FREE builds on CI's own MSYS2 CLANG64 toolchain were all flagged,
while the product binary scans clean on every platform.

So remove the stub and move self-update OUT of the process into install.ps1,
which runs while cbm is NOT running: Windows' image lock only blocks a
process from replacing ITSELF.  now prints the exact PowerShell
command (with the Unblock-File hint for Mark-of-the-Web); install.ps1 is
idempotent, so re-running it IS the update — it stops the daemon, renames the
running binary aside (the one mutation Windows permits on a running image),
publishes the new one, and sweeps retired copies.

Windows now matches Linux and macOS: ONE binary per platform.

  * packaging, install.ps1, npm and PyPI wrappers all carry a single binary
  * the launcher/payload ABI contract and ~2500 lines of stub state machinery
    are deleted
  * every daemon start, CLI call and hook fire loses a process spawn, a named
    pipe handshake and an stdio relay
  * test_windows_bundle_contract.sh is rewritten as an INVERTED contract: it
    now asserts no shipped surface can reintroduce a launcher/payload pair,
    and that install.ps1 retires the running binary before publishing

Verified: VirusTotal 0/67 on the packaged binary and 0/58 on install.ps1 (no
certificate involved); macOS and Linux full suites green; Windows guards all
green including the new update-handoff contract; npm 10/10; PyPI 3/3.

Signed-off-by: Martin Vogel <martin.vogel.tech@gmail.com>
2026-07-28 17:03:01 +02:00

247 lines
11 KiB
PowerShell

<#
.SYNOPSIS
Run the native-Windows product-surface test suite for codebase-memory-mcp.
.DESCRIPTION
Builds the product binary if it is not already present, stages it under its
release name, then runs the deterministic Windows integration tests under
tests/windows/ against it (real stdio / CLI / HTTP UI, real SQLite DB).
Windows ships ONE binary, exactly like Linux and macOS.
Two categories of test:
GUARDS - regression guards for Windows bugs already fixed on main.
They must stay GREEN (exit 0); a RED (exit 1) means the fix
regressed and fails this runner.
* test_non_ascii_path.py guards #636/#357 (fixed by #700)
* test_non_ascii_cache_dump.py guards #996 (writer cbm_fopen)
* test_hook_augment.py guards #618 (fixed by #619)
* test_ui_drive_listing.py guards #548 (roots field)
* test_cli_non_ascii_arg.py guards #423/#20 (wide-argv main())
* test_daemon_stability.py guards the daemon parameter
surface, crash recovery, busy-stop refusal, and churn
* test_windows_update_handoff.py guards that `update`
hands off to install.ps1 instead of replacing its own
running image (the removed launcher stub's only job)
KNOWN REDS - genuine, still-open Windows bugs reproduced at the product
surface. They are EXPECTED to be RED (exit 1) and are opt-in
(never gate CI). If one turns GREEN the underlying bug was
fixed and it should be promoted to a guard.
* (none currently - test_cli_non_ascii_arg.py was promoted to a
guard when the wide-argv fix for #423/#20 landed)
Indexing runs through the real supervisor -> worker spawn on every guard:
under the mandatory coordination daemon CBM_INDEX_SUPERVISOR=0 is a
fail-closed refusal seam, never an in-process fallback, so the old
determinism override would turn every indexing guard into a refusal.
On native Windows the MinGW/LLVM toolchain ships no libasan/libubsan, so the
build disables sanitizers (SANITIZE=). Where the toolchain provides
AddressSanitizer/UBSan (Linux containers, WSL), prefer scripts/test.sh.
.PARAMETER Binary
Path to an existing product executable. If omitted, the script builds it
(target selected by -Target) into build/c/.
.PARAMETER Target
Makefile.cbm target used when building: 'cbm-with-ui' (default; needed for the
drive-picker guard's embedded HTTP UI) or 'cbm' (no UI - the drive guard then
reports a precondition and is skipped).
.PARAMETER GuardsOnly
Run only the green guards (the CI gate). Skips the opt-in known-red repros.
.PARAMETER Make
Path to GNU make (default: 'make' on PATH; MSYS2 ships it at
C:\msys64\usr\bin\make.exe).
.EXAMPLE
pwsh -File scripts/test-windows.ps1
.EXAMPLE
pwsh -File scripts/test-windows.ps1 -GuardsOnly -Binary build\c\codebase-memory-mcp.exe
#>
[CmdletBinding()]
param(
[string]$Binary,
[ValidateSet("cbm-with-ui", "cbm")]
[string]$Target = "cbm-with-ui",
[switch]$GuardsOnly,
[string]$Make = "make"
)
$ErrorActionPreference = "Stop"
$repoRoot = Split-Path -Parent $PSScriptRoot
Set-Location $repoRoot
$python = (Get-Command python -ErrorAction SilentlyContinue)
if (-not $python) { $python = (Get-Command py -ErrorAction SilentlyContinue) }
if (-not $python) { throw "Python 3 is required to run the Windows tests." }
$py = $python.Source
# A writable Windows temp dir that GNU make forwards to the native gcc. MSYS2
# strips TMP/TEMP from the environment it hands native children, so pass them as
# make command-line variables (make exports those to recipe processes).
$tmp = $env:TEMP
if (-not $tmp) { $tmp = "$env:USERPROFILE\AppData\Local\Temp" }
function Resolve-Binary {
param([string]$Explicit)
if ($Explicit) { return (Resolve-Path $Explicit).Path }
$built = Join-Path $repoRoot "build\c\codebase-memory-mcp.exe"
if (Test-Path $built) { return $built }
Write-Host "Building $Target via Makefile.cbm ..." -ForegroundColor Cyan
& $Make "-j" "-f" "Makefile.cbm" $Target "SANITIZE=" "TMP=$tmp" "TEMP=$tmp" "TMPDIR=$tmp" | Out-Host
$buildExit = $LASTEXITCODE
if ($buildExit -ne 0) { throw "build failed (exit $buildExit)" }
if (-not (Test-Path $built)) { throw "binary not produced at $built" }
return $built
}
$bin = Resolve-Binary -Explicit $Binary
Write-Host "Binary: $bin" -ForegroundColor Green
$previousTemp = $env:TEMP
$previousTmp = $env:TMP
$previousTmpDir = $env:TMPDIR
$guardRoot = $null
try {
$userProfile = [Environment]::GetFolderPath([Environment+SpecialFolder]::UserProfile)
if (-not $userProfile) { throw "could not resolve the current user's profile directory" }
$guardRoot = Join-Path $userProfile ("cbm-windows-guards-root-" + [guid]::NewGuid().ToString("N"))
New-Item -ItemType Directory -Path $guardRoot | Out-Null
# GitHub-hosted runner profile children can inherit mutation-capable ACEs
# even though the profile ancestry itself passes the bounded trust policy. Replace that inheritance before creating any executable or
# Python temporary descendant. Use SIDs rather than localized account names.
$currentSid = [System.Security.Principal.WindowsIdentity]::GetCurrent().User
if (-not $currentSid) { throw "could not resolve the current user's SID" }
$guardAcl = [System.Security.AccessControl.DirectorySecurity]::new()
$guardAcl.SetOwner($currentSid)
$guardAcl.SetAccessRuleProtection($true, $false)
$guardRule = [System.Security.AccessControl.FileSystemAccessRule]::new(
$currentSid,
[System.Security.AccessControl.FileSystemRights]::FullControl,
([System.Security.AccessControl.InheritanceFlags]::ContainerInherit -bor
[System.Security.AccessControl.InheritanceFlags]::ObjectInherit),
[System.Security.AccessControl.PropagationFlags]::None,
[System.Security.AccessControl.AccessControlType]::Allow
)
$guardAcl.AddAccessRule($guardRule) | Out-Null
Set-Acl -LiteralPath $guardRoot -AclObject $guardAcl
$guardBundle = Join-Path $guardRoot ("cbm-windows-guards-" + [guid]::NewGuid().ToString("N"))
New-Item -ItemType Directory -Path $guardBundle | Out-Null
$guardBin = Join-Path $guardBundle "codebase-memory-mcp.exe"
Copy-Item -LiteralPath $bin -Destination $guardBin
# Ownership is never inherited on Windows: descendants created under the
# hardened root by an admin-group token can default to the Administrators
# SID, while the exe policy demands the exact current user as owner. Stamp
# the current SID explicitly on everything staged here.
foreach ($staged in @($guardBundle, $guardBin)) {
$stagedAcl = Get-Acl -LiteralPath $staged
$stagedAcl.SetOwner($currentSid)
Set-Acl -LiteralPath $staged -AclObject $stagedAcl
}
Write-Host "Guard bundle: $guardBin" -ForegroundColor Green
# The guards deliberately reject GitHub's shared D:\a ancestry and the
# hosted runner's inherited LocalAppData\Temp ACL. Keep staged fixtures and
# Python-created descendants below the accepted profile ancestry.
$env:TEMP = $guardRoot
$env:TMP = $guardRoot
$env:TMPDIR = $guardRoot
$env:PYTHONUTF8 = "1" # encode argv/stdio as UTF-8
# Green regression guards - must stay GREEN (exit 0). RED (exit 1) = the fix for
# the referenced issue regressed. The drive-picker guard needs the embedded HTTP
# UI (build target cbm-with-ui); against a non-UI binary it reports a precondition
# (exit 2) and is skipped rather than failed.
$guards = @(
"tests\windows\test_non_ascii_path.py",
"tests\windows\test_non_ascii_cache_dump.py",
"tests\windows\test_daemon_lifecycle.py",
"tests\windows\test_daemon_stability.py",
"tests\windows\test_hook_augment.py",
"tests\windows\test_ui_drive_listing.py",
"tests\windows\test_cli_non_ascii_arg.py",
"tests\windows\test_windows_update_handoff.py"
)
# Opt-in known-red repros - EXPECTED red (exit 1); never gate CI. Currently empty:
# test_cli_non_ascii_arg.py was promoted to a guard when #423/#20's wide-argv fix landed.
$knownReds = @()
$guardFailures = @()
$guardSkips = @()
$fixedKeepers = @()
Write-Host "`n--- Green guards ---" -ForegroundColor Cyan
foreach ($t in $guards) {
Write-Host "`n=== $t ===" -ForegroundColor Cyan
& $py $t $guardBin
$code = $LASTEXITCODE
if ($code -eq 0) {
Write-Host "GREEN ($t)" -ForegroundColor Green
} elseif ($code -eq 1 -or $t -eq "tests\windows\test_windows_update_handoff.py") {
Write-Host "RED ($t) - REGRESSION: a fixed Windows bug is broken again" -ForegroundColor Red
$guardFailures += $t
} elseif ($code -eq 2) {
# Exit 2 is the guards' DOCUMENTED precondition-skip contract; every
# other unexpected code (a crashed python, an access-violation status,
# a mistyped guard) is a FAILURE - an uncontracted exit once let a
# crashing guard read as an invisible skip under a green banner.
Write-Host "PRECONDITION ($t) exit=2 - skipped (see message above)" -ForegroundColor Yellow
$guardSkips += $t
} else {
Write-Host "FAILED ($t) exit=$code - crashed or exited outside the guard contract 0/1/2" -ForegroundColor Red
$guardFailures += $t
}
}
if (-not $GuardsOnly) {
Write-Host "`n--- Known reds (opt-in, expected red) ---" -ForegroundColor Cyan
foreach ($t in $knownReds) {
Write-Host "`n=== $t ===" -ForegroundColor Cyan
& $py $t $guardBin
$code = $LASTEXITCODE
if ($code -eq 1) {
Write-Host "RED ($t) - expected; the underlying Windows bug is still open" -ForegroundColor DarkYellow
} elseif ($code -eq 0) {
Write-Host "GREEN ($t) - the bug appears FIXED; promote this to a guard" -ForegroundColor Green
$fixedKeepers += $t
} else {
Write-Host "PRECONDITION ($t) exit=$code - skipped (see message above)" -ForegroundColor Yellow
}
}
}
} finally {
$env:TEMP = $previousTemp
$env:TMP = $previousTmp
$env:TMPDIR = $previousTmpDir
if ($guardRoot) {
Remove-Item -LiteralPath $guardRoot -Recurse -Force -ErrorAction SilentlyContinue
}
}
Write-Host ""
if ($guardSkips.Count -gt 0) {
Write-Host ("Guards skipped (precondition): {0} - e.g. the drive-picker guard " -f $guardSkips.Count) -ForegroundColor Yellow
Write-Host "needs a UI build (-Target cbm-with-ui, the default)." -ForegroundColor Yellow
}
if ($fixedKeepers.Count -gt 0) {
Write-Host ("Known-red repros that are now GREEN (promote to guards): {0}" -f ($fixedKeepers -join ", ")) -ForegroundColor Green
}
if ($guardSkips.Count -eq $guards.Count -and $guards.Count -gt 0) {
Write-Host "FAIL: every guard skipped - nothing was actually verified" -ForegroundColor Red
exit 1
}
if ($guardFailures.Count -gt 0) {
Write-Host ("REGRESSION: {0} green guard(s) went red: {1}" -f $guardFailures.Count, ($guardFailures -join ", ")) -ForegroundColor Red
Write-Host "A previously-fixed Windows bug is broken again (see the guard's docstring and its referenced issue)." -ForegroundColor Red
exit 1
}
Write-Host "All Windows green guards passed." -ForegroundColor Green
exit 0