test(e2e): centralize fake OpenAI-compatible server#5373
Merged
Conversation
Signed-off-by: Carlos Villela <cvillela@nvidia.com>
Contributor
|
Note Reviews pausedIt looks like this branch is under active development. To avoid overwhelming you with review comments due to an influx of new commits, CodeRabbit has automatically paused this review. You can configure this behavior by changing the Use the following commands to manage reviews:
Use the checkboxes below for quick actions:
No actionable comments were generated in the recent review. 🎉 ℹ️ Recent review info⚙️ Run configurationConfiguration used: Path: .coderabbit.yaml Review profile: CHILL Plan: Enterprise Run ID: 📒 Files selected for processing (1)
✅ Files skipped from review due to trivial changes (1)
📝 WalkthroughWalkthroughConsolidates a reusable fake OpenAI-compatible server (Node + TS fixture + shell helpers), migrates tests and E2E scripts to use it, updates PR-review advisor overlap/comment logic, adjusts nightly E2E provisioning to explicit Node setup, and applies formatting-only refactors across tools. ChangesE2E fake-server consolidation and advisor updates
Sequence Diagram(s)sequenceDiagram
participant TestCase as Test Case
participant Fixture as startFakeOpenAiCompatibleServer (fixture)
participant APIServer as fake-openai-compatible-api.mts (Node server)
participant RequestLog as requests.jsonl
TestCase->>Fixture: start({port, auth, chatContent, ...})
Fixture->>APIServer: spawn child process with FAKE_OPENAI_* env
Fixture->>APIServer: poll GET /v1/models until ready
APIServer-->>Fixture: READY + bound port
Fixture-->>TestCase: return baseUrl, requests(), close()
TestCase->>APIServer: POST /v1/chat/completions or /v1/responses
APIServer->>RequestLog: append JSONL request metadata
TestCase->>Fixture: fake.requests()
Fixture->>RequestLog: parse JSONL
Fixture-->>TestCase: array of typed requests
TestCase->>Fixture: fake.close()
Fixture->>APIServer: SIGTERM
APIServer-->>Fixture: exit
Fixture->>Fixture: cleanup temp files
Estimated code review effort🎯 3 (Moderate) | ⏱️ ~25 minutes Possibly related PRs
Suggested labels
Suggested reviewers
Poem
🚥 Pre-merge checks | ✅ 4 | ❌ 1❌ Failed checks (1 warning)
✅ Passed checks (4 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing Touches📝 Generate docstrings
🧪 Generate unit tests (beta)
Comment |
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
Contributor
There was a problem hiding this comment.
Actionable comments posted: 2
🧹 Nitpick comments (1)
tools/pr-review-advisor/analyze.mts (1)
773-808: ⚡ Quick winBound overlap file-fetch concurrency to avoid GitHub API burst throttling.
Line 773 launches up to 80 overlap checks concurrently, and each may call another paginated files API. This burst can hit secondary rate limits and silently under-report overlaps when the inner
catchfalls back tosameFiles = [].Suggested refactor
- const overlaps = await Promise.all( - openPulls - .filter((pull) => getPath<number>(pull, ["number"]) !== currentPrNumber) - .slice(0, 80) - .map(async (pull): Promise<OpenPrOverlap | null> => { + const overlaps: Array<OpenPrOverlap | null> = []; + for (const pull of openPulls + .filter((item) => getPath<number>(item, ["number"]) !== currentPrNumber) + .slice(0, 80)) { + overlaps.push( + await (async (): Promise<OpenPrOverlap | null> => { // existing body unchanged - }), - ); + })(), + ); + }🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the rest with a brief reason, keep changes minimal, and validate. In `@tools/pr-review-advisor/analyze.mts` around lines 773 - 808, The overlap checks currently run up to 80 concurrent tasks (the Promise.all over openPulls.slice(0, 80)) and each task may call githubRestPaginated, which can trigger GitHub secondary rate limits and cause sameFiles to silently fall back to empty; change the implementation to bound concurrency (e.g., use a concurrency limiter like p-limit or an explicit worker pool) when mapping openPulls to the overlap check so only a small number (e.g., 4–8) of githubRestPaginated calls run at once; update the code around overlaps/openPulls mapping (the async mapper that computes number, title, body, labels, linkedIssues, duplicateLinkedIssues, sameFiles) to schedule each mapper through the limiter and await Promise.all on the limited tasks instead of launching all 80 at once.
🤖 Prompt for all review comments with AI agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
Inline comments:
In `@test/e2e-scenario/fixtures/fake-openai-compatible.ts`:
- Around line 57-87: The readiness probe currently hardcodes "127.0.0.1" inside
canReachModels which ignores the configured host from
startFakeOpenAiCompatibleServer; modify canReachModels to accept a host
parameter and use that host instead of "127.0.0.1", update waitForReady to pass
the configured host through when calling canReachModels (thread the host from
wherever waitForReady is invoked, e.g., startFakeOpenAiCompatibleServer), and
ensure the http.get options use the provided host so readiness checks honor
non-loopback/IPv6 bindings.
In `@test/e2e/test-issue-2478-crash-loop-recovery.sh`:
- Around line 142-151: The preflight in
test/e2e/test-issue-2478-crash-loop-recovery.sh is incorrectly requiring python3
even though the compatible mock is started by node in
start_fake_openai_compatible_api (in
test/e2e/lib/openai-compatible-api-proof.sh); update the preflight checks to
verify the presence of node (and optionally that node supports the required
flag, e.g., by running `node --version` or a no-op with
`--experimental-strip-types`) instead of python3, leave the existing curl
readiness check intact, and ensure any environment-variable gating for
USE_COMPAT_MOCK=1 still triggers this node check.
---
Nitpick comments:
In `@tools/pr-review-advisor/analyze.mts`:
- Around line 773-808: The overlap checks currently run up to 80 concurrent
tasks (the Promise.all over openPulls.slice(0, 80)) and each task may call
githubRestPaginated, which can trigger GitHub secondary rate limits and cause
sameFiles to silently fall back to empty; change the implementation to bound
concurrency (e.g., use a concurrency limiter like p-limit or an explicit worker
pool) when mapping openPulls to the overlap check so only a small number (e.g.,
4–8) of githubRestPaginated calls run at once; update the code around
overlaps/openPulls mapping (the async mapper that computes number, title, body,
labels, linkedIssues, duplicateLinkedIssues, sameFiles) to schedule each mapper
through the limiter and await Promise.all on the limited tasks instead of
launching all 80 at once.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: Path: .coderabbit.yaml
Review profile: CHILL
Plan: Enterprise
Run ID: 1d82198e-a9aa-4f7a-9d34-8d5e98823ab3
📒 Files selected for processing (27)
biome.jsonscripts/generate-openclaw-config.mtsscripts/patch-openclaw-slack-deny-feedback.mtssrc/lib/messaging/applier/build/messaging-build-applier.mtstest/e2e-scenario/fixtures/fake-openai-compatible.tstest/e2e-scenario/live/double-onboard.test.tstest/e2e-scenario/live/token-rotation.test.tstest/e2e/lib/fake-openai-compatible-api.mtstest/e2e/lib/openai-compatible-api-proof.shtest/e2e/test-concurrent-gateway-ports.shtest/e2e/test-double-onboard.shtest/e2e/test-issue-2478-crash-loop-recovery.shtest/e2e/test-openshell-gateway-upgrade.shtools/advisors/git.mtstools/advisors/github.mtstools/advisors/io.mtstools/advisors/json.mtstools/advisors/session.mtstools/e2e-advisor/analyze.mtstools/e2e-advisor/comment.mtstools/e2e-advisor/dispatch.mtstools/e2e-advisor/scenario-comment.mtstools/e2e-advisor/scenarios.mtstools/e2e-scenarios/workflow-boundary.mtstools/pr-review-advisor/analyze.mtstools/pr-review-advisor/comment.mtstools/pr-review-advisor/workflow-boundary.mts
Signed-off-by: Carlos Villela <cvillela@nvidia.com>
Contributor
E2E Advisor RecommendationRequired E2E: None Full advisor summaryE2E Recommendation AdvisorFailed: Could not parse JSON from advisor output; see /home/runner/work/NemoClaw/NemoClaw/artifacts/e2e-advisor/e2e-advisor-raw-output.txt |
Contributor
Vitest E2E Scenario RecommendationRequired Vitest E2E scenarios: None Full Vitest E2E advisor summaryVitest E2E Scenario AdvisorFailed: Could not parse JSON from advisor output; see /home/runner/work/NemoClaw/NemoClaw/artifacts/e2e-advisor/e2e-scenario-advisor-raw-output.txt |
Contributor
PR Review AdvisorFindings: 0 needs attention, 1 worth checking, 0 nice ideas Review findings🛠️ Needs attention
🔎 Worth checking
🌱 Nice ideas
Consider writing more tests for
This is an automated advisory review. A human maintainer must make the final merge decision. |
Signed-off-by: Carlos Villela <cvillela@nvidia.com>
…patible-e2e # Conflicts: # tools/e2e-scenarios/workflow-boundary.mts
Signed-off-by: Carlos Villela <cvillela@nvidia.com>
Contributor
Selective E2E Results — ❌ Some jobs failedRun: 27454997874
|
Contributor
Selective E2E Results —
|
| Job | Result |
|---|---|
| concurrent-gateway-ports-e2e | |
| double-onboard-e2e |
…' into codex/fake-openai-compatible-e2e
Contributor
Vitest E2E Scenario Results — ✅ All jobs passedRun: 27454997804
|
Contributor
Selective E2E Results — ✅ All requested jobs passedRun: 27455188119
|
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Centralizes the generic OpenAI-compatible fake provider used by E2E tests so shell and Vitest scenarios share the same canned inference behavior. Also broadens Biome coverage to include
.mtshelpers and formats the existing.mtstool scripts.Changes
.mtsfake OpenAI-compatible API server with models, chat completions, responses, auth, streaming, and request logging support.**/*.mtsin Biome and apply formatting/linting to existing.mtsfiles.Type of Change
Verification
npx prek run --all-filespassesnpm testpassesnpm run docsbuilds without warnings (doc changes only)Signed-off-by: Carlos Villela cvillela@nvidia.com
Summary by CodeRabbit
Chores
Tests
CI