Implement Stage A: rule engine, validation, audit, and safety gates

Completes 109 of 121 tasks. Every remaining task needs a tenant connection
(T055, T056, T101-T103) or an Azure Automation account (T115-T121).

  354 offline Pester tests      PASS
  Engine purity (Principle IV)  PASS
  Sanitization (SC-013)         PASS  (156 files)
  Graph module loaded in tests  none  (SC-008 holds)

What landed
  - Four-layer configuration validation with stable finding codes, covering
    every VR-002 and VR-003 condition, plus a 23-fixture invalid-config corpus
  - Run loop, audit records (NDJSON through a single sink), summaries,
    reconciliation, and exit codes 0-6
  - Persistence behind a single write-body builder whose result always has
    exactly one key
  - Invoke-PersonaEngine.ps1 and Edit-PersonaEngineConfig.ps1
  - Six docs, two pipelines, traceability matrix, V-5a and sanitization records

Three deviations from tasks.md, each recorded in its status block

  T033 is not in Resolve-UserPersona. evaluationErrorThreshold is run-level
  state and the rule engine is pure; a counter there would break Principle IV.
  It lives in New-PersonaRunCounter and is applied in the run loop.

  A new src/Engine/ layer holds Invoke-PersonaEngineRun. The entry script
  imports the manifest, which requires Microsoft.Graph.Authentication, so a
  loop living only inside it could not run on a machine without the Graph SDK
  and SC-004 could not be proven at all. The entry script is now a thin
  wrapper and what ships is what is tested.

  The invalid-config corpus is generated by a committed script, with the
  generated fixtures committed too, so a reviewer sees the fixture in the diff.

Defects found by running the code, not by reading it

  Group and role ID lists were double-wrapped: @(Get-PersonaGroupIdPage ...)
  around a comma-returned array collapsed every membership list into one
  bogus space-joined entry. That is a silent false non-match, exactly what
  FR-013 exists to prevent.

  A 403 whose status appears only in the exception message parsed as $null,
  which the retry policy treats as a transport error - five requests per
  account against a tenant already refusing. Status extraction now falls back
  to the message text, bounded to 400-599.

  The sanitization scan walked tracked files only, so it covered 34 of 156
  files and none of this phase's code. It now scans untracked non-ignored
  files too, and a negative control confirms it catches a planted leak.

  Test-Json reports one error per violating location, not first-failure-only
  as the V-5a draft claimed. Record and pin corrected.

Enforcement remains blocked on the V-4 security sign-off.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
This commit is contained in:
2026-08-20 21:48:19 -04:00
parent c59c85dd55
commit cdc6bb33d3
124 changed files with 16638 additions and 199 deletions
+111
View File
@@ -0,0 +1,111 @@
#Requires -Modules @{ ModuleName = 'Pester'; ModuleVersion = '5.0.0' }
<#
NFR-003 and the OTD-007 retry table.
The non-retryable cases carry the operational weight. A 403 retried five times per
account turns an instant authorization failure into a long, expensive one, and
hammers a tenant that is already refusing - so status extraction has to work even
when the code appears only in the exception message, which is how
Invoke-MgGraphRequest reports several of its failures.
#>
BeforeAll {
$repoRoot = Split-Path (Split-Path $PSScriptRoot -Parent) -Parent
. (Join-Path $repoRoot 'tests/TestHelpers.ps1')
foreach ($file in (Get-PersonaSourceFile -RepoRoot $repoRoot)) { . $file }
function New-StatusError {
param([int] $Status, [string] $Reason = 'Failed')
[System.Management.Automation.ErrorRecord]::new(
[System.Exception]::new("Response status code does not indicate success: $Status ($Reason)."),
'GraphError', [System.Management.Automation.ErrorCategory]::InvalidResult, $null)
}
}
Describe 'Status extraction from an error record' {
It 'reads a status embedded only in the message' {
Get-PersonaGraphStatusCode -ErrorRecord (New-StatusError -Status 403 -Reason 'Forbidden') | Should -Be 403
}
It 'reads each status the retry table names' {
foreach ($status in @(400, 401, 403, 404, 409, 429, 500, 502, 503, 504)) {
Get-PersonaGraphStatusCode -ErrorRecord (New-StatusError -Status $status) | Should -Be $status
}
}
It 'returns null for a transport failure with no status anywhere' {
$record = [System.Management.Automation.ErrorRecord]::new(
[System.Exception]::new('The operation was canceled.'),
'Timeout', [System.Management.Automation.ErrorCategory]::OperationTimeout, $null)
Get-PersonaGraphStatusCode -ErrorRecord $record | Should -BeNullOrEmpty
}
It 'does not mistake an unrelated three-digit number for a status' {
$record = [System.Management.Automation.ErrorRecord]::new(
[System.Exception]::new('Processed 250 objects before the connection dropped.'),
'Transport', [System.Management.Automation.ErrorCategory]::ConnectionError, $null)
# 250 is outside the HTTP error range, so it is not treated as a status and
# the failure stays retryable - which is the correct answer for a dropped
# connection.
Get-PersonaGraphStatusCode -ErrorRecord $record | Should -BeNullOrEmpty
}
}
Describe 'Retry policy (OTD-007)' {
# One It per status rather than a loop: Should -Invoke counts across the whole It
# block, so a loop would accumulate calls and the second iteration would fail on
# the first one's arithmetic.
It 'never retries status <_>' -ForEach @(400, 401, 403, 404, 409) {
$status = $_
Mock Invoke-MgGraphRequest { throw "Response status code does not indicate success: $status (Failed)." }.GetNewClosure()
{ Invoke-PersonaGraphRequest -Uri '/v1.0/users' -BaseDelayMs 1 } | Should -Throw
Should -Invoke Invoke-MgGraphRequest -Times 1 -Exactly -Because "status $status is a client-side defect that retrying would only hide"
}
It 'retries a retryable status up to the attempt limit' {
Mock Invoke-MgGraphRequest { throw 'Response status code does not indicate success: 503 (Service Unavailable).' }
{ Invoke-PersonaGraphRequest -Uri '/v1.0/users' -MaxAttempts 3 -BaseDelayMs 1 } | Should -Throw
Should -Invoke Invoke-MgGraphRequest -Times 3 -Exactly
}
It 'retries a transport failure that carries no status' {
Mock Invoke-MgGraphRequest { throw 'The operation was canceled.' }
{ Invoke-PersonaGraphRequest -Uri '/v1.0/users' -MaxAttempts 2 -BaseDelayMs 1 } | Should -Throw
Should -Invoke Invoke-MgGraphRequest -Times 2 -Exactly
}
It 'returns as soon as an attempt succeeds' {
$script:attempts = 0
Mock Invoke-MgGraphRequest {
$script:attempts++
if ($script:attempts -lt 3) { throw 'Response status code does not indicate success: 429 (Too Many Requests).' }
@{ value = @('ok') }
}
$result = Invoke-PersonaGraphRequest -Uri '/v1.0/users' -BaseDelayMs 1
$result['value'] | Should -Be @('ok')
Should -Invoke Invoke-MgGraphRequest -Times 3 -Exactly
}
It 'passes the body through unchanged as JSON' {
# SC-005 depends on this: the body a test captures must be the body sent.
$script:sentBody = $null
Mock Invoke-MgGraphRequest { $script:sentBody = $Body; @{} }
$null = Invoke-PersonaGraphRequest -Uri '/v1.0/users/x' -Method 'PATCH' -Body @{ 'extension_x_Persona' = 'Employee' }
($script:sentBody | ConvertFrom-Json).'extension_x_Persona' | Should -Be 'Employee'
}
}