Repository navigation
Preserve live Railway settings and gate deployments on tested commits - #130
jackshim415 merged 10 commits into
Conversation
Preserve current staging sources, sleeping policy and inherited preview API limits. Record named-partial adoption, redacted rollback snapshots and empty production/staging/preview plans without applying changes. Refs: LVT-336
…rship Refs: LVT-336
Codex Review Summary: ✅ CompletedThis comment shows the latest Codex review activity on this pull request.
ℹ️ About Codex in GitHubYour team has set up Codex to review pull requests in this repo. Reviews are triggered when you
Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings. |
|
🚅 Deployed to the lovat-pr-130 environment in lovat
1 service not affected by this PR
|
Lovat AI reviewStatus: completed · Model: Reviewed commit: Critical: 0 · High: 0 · Medium: 0 · Low: 0 Overall assessmentThis batch is mostly CI wiring, Railway IaC test updates, and read-only audit evidence. The CodeQL workflow conversion to workflow_call with an aggregate result gate is internally consistent (skipped analysis passes, failed/cancelled selected analysis fails). The web-images filter mirrors the removed pull_request paths, and the config tests correctly encode the preserved live settings (sleeping on non-worker services, main branch for both environments, env-gated checkSuites). The evidence JSON contains only resource IDs and variable names, no secret values. No concrete defect with an explainable failure mode was found in the changed lines. This batch aligns the Railway IaC graph with the observed live configuration documented in the evidence files: the staging source branch moves from the stale offseason branch to main, sleeping is declared for the four user-facing apps (worker correctly excluded), and the API container limit now covers previews as well as staging, matching the recorded audit. The CI gate changes (mandatory codeql result, web-images selector) are internally consistent and covered by updated unit tests. The CodeQL matrix change intentionally replaces the workflow force-all with the Actions analysis and is explicitly tested. No concrete defect with an explainable failure mode is present in the changed lines. No high-confidence issues were found in the reviewed changes. Review details
Limitations
|
📱 Connect your app to this PR deploymentScan a QR code to use this PR's Server API. Worker changes use the same API; Server URL: |
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 11732690a1
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
| | ------------ | -------------------------------------- | --------------------------------------------------------- | --------------------------- | ------------ | | ||
| | production | `9f0088a8-9fe7-4afa-b634-cd1db046f045` | `main` | `lovat-apps`, all five apps | No changes | | ||
| | staging | `1902828c-65ae-455c-90c8-5a370b7cedab` | `main` | `lovat-apps`, all five apps | No changes | | ||
| | lovat-pr-128 | `d42d9835-c01a-48c4-bcd2-424fa61e93c3` | `jackshim415/lvt-334-server-redact-request-analytics-and` | `lovat-apps`, all five apps | No changes | |
There was a problem hiding this comment.
Mark preview partial ownership as unverified
The preview row still reports that lovat-apps owns all five applications, but preview128-plan-summary.json records an empty ownership object and the following paragraph explicitly says that PR #128 has not been adopted into the partial. This contradictory cutover evidence can lead operators to treat preview ownership as complete, so the table should identify it as unowned or pending adoption.
Useful? React with 👍 / 👎.
…nfrastructure-verify-lovat-railway-iac # Conflicts: # .railway/README.md # .railway/config.test.mjs
Refs: LVT-336, LVT-337
Reuse the four SDK 58 patch updates already verified on the Collection native-testing branch. Refs: LVT-337, LVT-274
Refs: LVT-334
Refs: LVT-336, LVT-337
…nfrastructure-verify-lovat-railway-iac
Refs: LVT-337
Keep Railway’s observed source branches, sleeping settings, and preview limits in the named app partial. The latest production dry run proposes only the five
source.checkSuitesupdates. The preview dry run also sets the Worker’s currentcontinuousdefault explicitly. The production and preview plans reported zero diagnostics. Preview partial ownership and credential isolation still need separate evidence.Put applicable web-image HTTP smoke tests and CodeQL analysis under the required
Monorepo CIstatus. CodeQL allows a verified empty selection and fails the gate when a selected analysis is missing, fails, or is cancelled.RAILWAY_WAIT_FOR_CI=truerenders the provider change for review. The runbook explains how to verify a passing and a failing deployment by commit SHA.Verification
.railwaytypecheck, all nine configuration tests,actionlint, all 18 CI selection/result/CodeQL matrix tests, andgit diff --checkpassed.expo install --check, TypeScript, lint, formatting, tests, and the existing high-severity audit threshold.Refs: LVT-336, LVT-337.