Set shortLivedArtifactsRetentionDays to 3

[aholstrup1]

What & why

Increases shortLivedArtifactsRetentionDays from the default of 1 to 3 in .github/AL-Go-Settings.json.

The default 1-day retention on the short-lived Apps/TestApps build artifacts is shorter than a realistic re-run window. When a PR run is re-run more than a day after its build jobs produced those artifacts, the app binaries have already expired while the long-lived (30-day) BuildOutput metadata survives. AL-Go's dependency resolver then finds no current-run app artifacts locally and silently falls back to baseline main branch apps, so tests execute against stale baseline binaries and cannot catch newly added or changed tests. This produced a false-green run (Shopify test UnitTestCalcPriceUsesCurrentWorkDate from #9404 never ran in BCApps CI but failed downstream in NAV). Widening the window to 3 days makes the re-run case work correctly.

Linked work

Fixes AB#642371

How I validated this

• I read the full diff and it contains only changes I intended.
• I built the affected app(s) locally with no new analyzer warnings.
• I ran the change in Business Central and confirmed it behaves as expected.
• I added or updated tests for the new behavior, or explained below why none are needed.

What I tested and the outcome

Settings-only change to CI artifact retention; no app code affected and no tests apply. Verified the JSON is valid and the new key is placed alongside the existing incremental-build settings.

Risk & compatibility

Low risk. Only affects retention duration of short-lived CI artifacts (slightly higher storage usage). No product code, upgrade, or data impact.

[github-actions]

Copilot PR Review

Iteration 1 · Outcome: completed

The PR changes only .github/AL-Go-Settings.json (adds "shortLivedArtifactsRetentionDays": 3), a CI/build configuration file. No AL source files are changed, so no AL object, declaration, or code-convention signal exists for any sub-skill's Worklist step to match. Every sub-skill's worklist was empty and each returned outcome: completed with zero findings. The self-review pass found no cross-cutting AL/BC defect either; the added property is present in the referenced AL-Go settings.schema.json and uses a plausible value, so no agent finding was warranted.

Knowledge source: https://github.com/microsoft/BCQuality@186d8a131465475c79244d994acb872cd5c0d4bf

Orchestrator pre-filter (2 file(s) excluded)

• layer-disabled (knowledge) : 2 file(s)

_{Findings produced by the AL review agent v1.7.3. Reply 👎 on any inline comment to flag false positives.}