test(e2e): avoid inference calls in hosted reachability probe

[cv]

Summary

Changes the hosted inference E2E preflight from an authenticated chat completion to a lightweight HTTP reachability probe. This avoids spending model requests and tripping HTTP 429 during full parallel nightly runs before the scenario reaches the behavior under test.

Related Issue

Related to #5406 nightly follow-up.

Changes

• Update test/e2e/lib/ci-compatible-inference.sh so nemoclaw_e2e_probe_hosted_inference checks network/TLS reachability only and accepts any non-000 HTTP response.
• Add support-test coverage proving the probe does not call /chat/completions.

Type of Change

• Code change (feature, bug fix, or refactor)
• Code change with doc updates
• Doc only (prose changes, no code sample modifications)
• Doc only (includes code sample changes)

Verification

• Git hooks passed during commit and push, or npx prek run --from-ref main --to-ref HEAD passes
• Targeted tests pass for changed behavior
• Full npm test passes (broad runtime changes only)
• Tests added or updated for new or changed behavior
• No secrets, API keys, or credentials committed
• Docs updated for user-facing behavior changes
• npm run docs builds without warnings (doc changes only)
• Doc pages follow the style guide (doc changes only)
• New doc pages include SPDX header and frontmatter (new pages only)

Targeted verification:

• npx biome check --write test/e2e-scenario/support-tests/hosted-inference.test.ts
• bash -n test/e2e/lib/ci-compatible-inference.sh
• npx vitest run --project cli test/e2e-scenario/support-tests/hosted-inference.test.ts

Docs review: no user-facing docs changes needed; this is E2E harness stabilization only.

Note: local broad hooks still fail in unrelated runtime recovery preload tests because temp preload files are seen as group-writable (mode=664), matching prior local hook failures. Targeted changed tests passed.

---

Signed-off-by: Carlos Villela cvillela@nvidia.com

Summary by CodeRabbit

• Tests

  • Added end-to-end coverage for hosted inference probe validation, including lightweight reachability assertions and negative cases when connectivity fails or the probe reports errors.

• Refactor

  • Updated the hosted inference probe to run a lightweight network/TLS connectivity check only, avoiding authenticated requests and preventing inference/auth-related endpoints from being called.

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Enterprise

Run ID: d6aa061a-ea4a-432d-b4cf-ebf6f55578b7

📥 Commits

Reviewing files that changed from the base of the PR and between 5910ba0 and cfc7a3b.

📒 Files selected for processing (2)

• test/e2e-scenario/support-tests/hosted-inference.test.ts
• test/e2e/lib/ci-compatible-inference.sh

🚧 Files skipped from review as they are similar to previous changes (1)

• test/e2e-scenario/support-tests/hosted-inference.test.ts

---

📝 Walkthrough

Walkthrough

nemoclaw_e2e_probe_hosted_inference in ci-compatible-inference.sh is simplified from an authenticated probe (issuing /chat/completions or /models requests with an API key) to a single curl call that checks network/TLS reachability by inspecting the HTTP status of the hosted inference base_url. A new E2E test validates this behavior by stubbing curl and asserting the probe does not invoke chat/completions or /models.

Changes

Hosted-inference probe simplification and test coverage

Layer / File(s)	Summary
Probe function reduced to curl reachability check test/e2e/lib/ci-compatible-inference.sh	nemoclaw_e2e_probe_hosted_inference now only curls base_url, captures the HTTP status, and returns success when the status is present and not 000; all authenticated inference and /models request logic is removed.
E2E test suite verifying probe skips authenticated requests test/e2e-scenario/support-tests/hosted-inference.test.ts	Adds Node.js imports, a COMPAT_HELPER path constant, a runHostedProbe helper that stubs curl in a temp directory, and E2E tests that verify the probe runs successfully while avoiding chat/completions and /models requests; includes negative cases confirming failures on HTTP status 000 or nonzero exit codes.

Estimated code review effort

🎯 2 (Simple) | ⏱️ ~12 minutes

Possibly related PRs

• NVIDIA/NemoClaw#5395: Both PRs adjust the hosted-inference CI probe behavior — this PR removes the chat/completions request from the probe, while that PR defaults NEMOCLAW_PREFERRED_API to openai-completions to avoid stricter chat/response constraints in the same hosted-inference flow.

Poem

🐇 A curl to the sky, just checking the door,
No tokens to send, no model to implore.
The probe hops lightly, status in paw,
"Not 000!" it cheers — that's the only law.
The test stubs the curl, the log tells the tale:
No chat/completions — the rabbit won't fail! 🎉

🚥 Pre-merge checks | ✅ 4 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Docstring Coverage	⚠️ Warning	Docstring coverage is 0.00% which is insufficient. The required threshold is 80.00%.	Write docstrings for the functions missing them to satisfy the coverage threshold.

✅ Passed checks (4 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The PR title clearly and specifically summarizes the main change: avoiding inference calls in the hosted reachability probe.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

📝 Generate docstrings

• Create stacked PR
• Commit on current branch

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch fix/e2e-hosted-inference-reachability-probe

---

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[github-code-quality]

Code Coverage Overview

Languages: TypeScript

TypeScript / code-coverage/plugin

The overall coverage in the fix/e2e-hosted-infer... branch is 96%. Coverage data for the main branch is not yet available.

Show a code coverage summary of the most covered files.

File	main	fix/e2e-hosted-infer... cfc7a3b	+/-
nemoclaw/src/se...cret-scanner.ts	—	100%	—
nemoclaw/src/commands/slash.ts	—	100%	—
nemoclaw/src/li...bprocess-env.ts	—	100%	—
nemoclaw/src/bl...eprint/state.ts	—	98%	—
nemoclaw/src/onboard/config.ts	—	98%	—
nemoclaw/src/bl...int/snapshot.ts	—	97%	—
nemoclaw/src/bl...print/runner.ts	—	95%	—
nemoclaw/src/co...ration-state.ts	—	94%	—
nemoclaw/src/bl...ate-networks.ts	—	94%	—
nemoclaw/src/index.ts	—	94%	—

TypeScript / code-coverage/cli

The overall coverage in the fix/e2e-hosted-infer... branch is 44%. Coverage data for the main branch is not yet available.

Show a code coverage summary of the most covered files.

File	main	fix/e2e-hosted-infer... cfc7a3b	+/-
src/lib/state/o...oard-session.ts	—	90%	—
src/lib/inference/local.ts	—	77%	—
src/lib/sandbox/config.ts	—	72%	—
src/lib/inference/nim.ts	—	72%	—
src/lib/onboard/preflight.ts	—	64%	—
src/lib/state/sandbox.ts	—	55%	—
src/lib/onboard...er-gpu-patch.ts	—	50%	—
src/lib/actions...licy-channel.ts	—	49%	—
src/lib/policy/index.ts	—	48%	—
src/lib/onboard.ts	—	17%	—

---

_{Updated June 14, 2026 02:36 UTC
Code Coverage is in Public Preview. Learn more and provide us with your feedback.}

[github-actions]

E2E Advisor Recommendation

Required E2E: None
Optional E2E: cloud-onboard-e2e, cloud-e2e, cloud-inference-e2e

Workflow run

Full advisor summary

E2E Recommendation Advisor

Base: origin/main
Head: HEAD
Confidence: high

Required E2E

• None.

Optional E2E

• cloud-onboard-e2e (medium): Best targeted confidence check for the changed hosted inference preflight: this script sources ci-compatible-inference.sh, calls nemoclaw_e2e_probe_hosted_inference before install/onboard, and then validates hosted compatible onboarding.
• cloud-e2e (medium): Broader optional smoke for the full install/onboard/sandbox flow using the same hosted inference helper and preflight path.
• cloud-inference-e2e (medium): Adjacent optional coverage for the hosted compatible inference environment configured by the shared helper, with an actual inference.local chat request after sandbox setup.

New E2E recommendations

• None.

[github-actions]

Vitest E2E Scenario Recommendation

Required Vitest E2E scenarios: None
Optional Vitest E2E scenarios: None

Workflow run

Full Vitest E2E advisor summary

Vitest E2E Scenario Advisor

Base: origin/main
Head: HEAD
Confidence: high

Required Vitest E2E scenarios

• None. No Vitest live scenario dispatch is required. The PR changes a support-test plus the legacy shell helper test/e2e/lib/ci-compatible-inference.sh; the live Vitest scenario workflow runs registry-scenarios.test.ts and wired test/e2e-scenario/live/*.test.ts jobs, and none of those dispatches directly exercises nemoclaw_e2e_probe_hosted_inference from this shell helper.

Optional Vitest E2E scenarios

• None.

Relevant changed files

• test/e2e-scenario/support-tests/hosted-inference.test.ts

[github-actions]

PR Review Advisor

Findings: 0 needs attention, 0 worth checking, 0 nice ideas
Since last review: 3 prior items resolved, 0 still apply, 0 new items found

Workflow run details

This is an automated advisory review. A human maintainer must make the final merge decision.