test(bdd): invoke an LLM function in the single-cluster helmfile feature

[along-2017]

Why

The single-cluster Helmfile feature installs the LLM gateway and
router but never routes a request through them. A regression that
leaves the releases deployed while the serve path is broken passes
the suite. This adds the first LLM-type function coverage.

What changed

• tests/bdd/features/single-cluster-helmfile.feature: new
  @llm-function-type scenario after the echo function lifecycle. It
  creates a function of type LLM backed by the OpenAI-compatible
  sample image, deploys it, invokes a chat completion through the LLM
  gateway asserting the response envelope, and asserts an
  unauthenticated gateway request returns 401. Existing DSL steps
  only.
• tests/bdd/godog_test.go: wiring test seeds canned results for the
  LLM invoke and the curl check, and asserts the LLM create, invoke,
  and no-auth commands ran.

Customer Release Notes

Not customer visible.

Plan Summary

Not applicable

Usage

Not applicable

Testing

• tests/bdd: go test -short ./... green (wiring test executes the new
  scenario against the fake runner); go vet green.
• Live TestSingleClusterHelmfile not yet run: the scenario references
  nvcf-openai-compatible-sample:1.0.0 under the sample org/team, which
  must be published as a multi-arch tag first. The live run happens
  once the image exists; the invoke content assertion is tightened
  then.

Notes

The image tag 1.0.0 is a placeholder for the first published
multi-arch tag; update the scenario if the published tag differs.

References

Relates to #1019

Related Pull Requests

None

Dependencies

None

Summary by CodeRabbit

• Tests
  • Added end-to-end coverage for the LLM function lifecycle, including creation, deployment, API-key generation, and synchronous OpenAI-compatible inference.
  • Added response validation to confirm successful invocation behavior.
  • Added coverage verifying that unauthenticated gateway requests are rejected with HTTP 401.
  • Expanded deployment wiring checks for model and runtime configuration.

[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: 027bf123-173d-462e-b6a9-ffdabc18cf37

📥 Commits

Reviewing files that changed from the base of the PR and between 171757e and da58706.

📒 Files selected for processing (2)

• tests/bdd/features/single-cluster-helmfile.feature
• tests/bdd/godog_test.go

Included review availability: Your plan provides up to 12 included reviews per hour; 11 remain after this review.

---

📝 Walkthrough

Walkthrough

The BDD tests add an OpenAI-compatible LLM function lifecycle. They verify deployment configuration, API-key generation, synchronous invocation, response validation, and HTTP 401 rejection for unauthenticated gateway requests.

Changes

LLM gateway testing

Layer / File(s)	Summary
LLM function lifecycle and authentication tests/bdd/features/single-cluster-helmfile.feature, tests/bdd/godog_test.go	The scenario creates and deploys an LLM function with the expected image, function type, and model. Fake-runner responses support authenticated chat-completions invocation and unauthenticated HTTP 401 validation.

Estimated code review effort: 2 (Simple) | ~10 minutes

Merge Risk: 🔵 Low · up to da587

The PR adds coverage for LLM invocation and gateway authentication without changing product runtime code. The live scenario depends on publishing and pulling the sample image first, so it is mergeable with explicit owner follow-up to verify that dependency before live execution.

Sequence Diagram(s)

sequenceDiagram
  participant BDDScenario
  participant FakeRunner
  participant LLMGateway
  BDDScenario->>FakeRunner: create and deploy LLM function
  BDDScenario->>LLMGateway: invoke chat-completions endpoint with API key
  LLMGateway-->>BDDScenario: synchronous OpenAI-compatible response
  BDDScenario->>LLMGateway: invoke gateway without API key
  LLMGateway-->>BDDScenario: HTTP 401

Loading

Suggested reviewers: nvaghela-oss

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title follows Conventional Commits with the test(bdd) type and accurately describes the added BDD coverage for invoking an LLM function.
Docstring Coverage	✅ Passed	Docstring coverage is 100.00% which is sufficient. The required threshold is 80.00%. Docstring coverage is scoped to functions touched by this diff. Analyzed 1 functions across 1 files. (1 skipped: 1 unsupported.)
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.

✨ 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 test/bdd/llm-function-single-cluster

---

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