chore(inference): remove retired Build model paths

[ericksoa]

Summary

Remove obsolete NVIDIA Build compatibility paths for MiniMax M2.7 and Qwen 3.5 397B.
NVIDIA Endpoints onboarding keeps its current M3 choices, while independent Hermes Provider entries remain unchanged.

Related Issue

Fixes #7645

Changes

• Remove the obsolete minimaxai/minimax-m2.7 featured-feed ID and label rewrite.
• Remove qwen/qwen3.5-397b-a17b from the extended NVIDIA Endpoints validation budget.
• Replace retired Build model IDs in unrelated tests with current or provider-neutral fixture values.
• Preserve minimax/minimax-m2.7 and stepfun/step-3.5-flash in the independent Hermes Provider catalog.

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)

Quality Gates

• Tests added or updated for changed behavior
• Existing tests cover changed behavior — justification:
• Tests not applicable — justification:
• Docs updated for user-facing behavior changes
• Docs not applicable — justification: Published NVIDIA Endpoints pages already list M3, omit Qwen 3.5 397B, and describe Hermes Provider separately.
• Sensitive paths changed (security, policy, credentials, preflight, onboarding, inference, runner, sandbox, or messaging)
• Sensitive-path review completed or maintainer-approved waiver recorded — reviewer/approval link/justification: The reviewed diff only removes model-specific normalization and timeout entries. Authentication, credentials, provider routing, and manual /models validation remain unchanged.
• Non-success, skipped, or missing CI check accepted by maintainer — check name, approval link, and follow-up issue:

Documentation Writer Review

• Documentation writer subagent reviewed the completed changes
• Result: no-docs-needed
• Evidence: docs/inference/choose-model.mdx and docs/inference/use-nvidia-endpoints.mdx already list M3 and omit Qwen 3.5 397B. Hermes Provider documentation already describes its independent catalog.
• Agent: Codex Desktop

DGX Station Hardware Evidence

• Tested on DGX Station
• Tested commit:
• Station profile/scenario:
• Result:
• Supporting evidence:

Verification

• PR description includes a Signed-off-by: line and every commit appears as Verified in GitHub
• Normal pre-commit, commit-msg, and pre-push hooks passed, or npm run check:diff passed when hooks were skipped or unavailable
• Targeted behavior tests pass for the current change set, or tests are marked not applicable above — command/result or justification: CLI affected tests passed with 212 tests and 1 skip. test/onboard-selection.test.ts passed 64 tests. The Hermes workflow-boundary test passed 18 tests.
• Applicable broad gate passed — npm test for broad runtime/test-harness changes; npm run check for repo-wide validation/coverage changes — command/result:
• Quality Gates section completed with required justifications or waivers
• No secrets, API keys, or credentials committed
• 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)

---

Signed-off-by: Aaron Erickson aerickson@nvidia.com

Summary by CodeRabbit

• New Features

  • NVIDIA featured-model listings now preserve the model IDs and labels supplied by the catalog.
  • Extended validation timing is applied to DeepSeek V4 Flash for more reliable endpoint checks.

• Bug Fixes

  • Updated inference routing, sandbox status, onboarding, and endpoint checks to reflect current model catalog entries.
  • Improved model-selection and workflow validation coverage for updated NVIDIA model configurations.

[Copilot]

Copilot was unable to review this pull request because the user who requested the review has reached their quota limit.

[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: 4f7cced4-bc97-4b01-8653-ef3597037796

📥 Commits

Reviewing files that changed from the base of the PR and between a8748ab and eda3ede.

📒 Files selected for processing (10)

• src/lib/inference/config.test.ts
• src/lib/inference/health.test.ts
• src/lib/inference/nvidia-featured-models.test.ts
• src/lib/inference/nvidia-featured-models.ts
• src/lib/inference/onboard-probes.test.ts
• src/lib/inference/onboard-probes.ts
• src/lib/inventory/index.test.ts
• src/lib/onboard/compatible-endpoint-smoke.test.ts
• test/e2e/support/hermes-workflow-boundary.test.ts
• test/onboard-selection.test.ts

💤 Files with no reviewable changes (2)

• src/lib/inference/nvidia-featured-models.test.ts
• src/lib/inference/health.test.ts

---

📝 Walkthrough

Walkthrough

Changes

NVIDIA Build compatibility cleanup

Layer / File(s)	Summary
Featured-model catalog normalization src/lib/inference/nvidia-featured-models.ts, src/lib/inference/nvidia-featured-models.test.ts, test/onboard-selection.test.ts	Removes the Minimax M2.7-to-M3 catalog rewrite and derives labels directly from sanitized catalog text; the selection fixture now uses Minimax M3.
NVIDIA endpoint validation coverage src/lib/onboard/onboard-probes.ts, src/lib/inference/health.test.ts, src/lib/inference/onboard-probes.test.ts, src/lib/inference/config.test.ts	Removes Qwen 3.5 397B from extended validation and timeout cases while retaining DeepSeek V4 Flash coverage and updating NVIDIA routing assertions to Nemotron.
Provider-neutral test fixtures src/lib/inventory/index.test.ts, src/lib/onboard/compatible-endpoint-smoke.test.ts, test/e2e/support/hermes-workflow-boundary.test.ts	Replaces retired model identifiers in status, smoke-test, and workflow-boundary fixtures with provider-neutral values.

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

Suggested labels: bug-fix

Suggested reviewers: cv, prekshivyas

🚥 Pre-merge checks | ✅ 4 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Docstring Coverage	⚠️ Warning	Docstring coverage is 50.00% 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 title is concise and accurately summarizes the main change: removing retired Build model paths.
Linked Issues check	✅ Passed	The changes remove the Minimax rewrite and Qwen validation path, while tests switch to current or neutral model IDs as requested.
Out of Scope Changes check	✅ Passed	The modified files stay focused on removing retired model paths and updating related tests, with no clear unrelated code changes.

✨ 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/7645-retired-build-model-cleanup

---

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

[github-code-quality]

Code Coverage Overview

Languages: TypeScript

TypeScript / code-coverage/plugin

The overall coverage in commit eda3ede in the fix/7645-retired-bui... branch remains at 96%, unchanged from commit a8748ab in the main branch.

TypeScript / code-coverage/cli

The overall coverage in commit eda3ede in the fix/7645-retired-bui... branch remains at 80%, unchanged from commit a8748ab in the main branch.

Show a code coverage summary of the most impacted files.

File	main a8748ab	fix/7645-retired-bui... eda3ede	+/-
src/lib/actions...time-command.ts	100%	82%	-18%
src/lib/actions...-add-restart.ts	19%	10%	-9%
src/lib/actions...on-readiness.ts	100%	91%	-9%
src/lib/actions...lution-probe.ts	95%	88%	-7%
src/lib/actions...x/mcp-bridge.ts	41%	35%	-6%
src/lib/actions...e-validation.ts	84%	81%	-3%
src/lib/onboard...shboard-port.ts	93%	90%	-3%
src/lib/actions...dbox/destroy.ts	95%	93%	-2%
src/lib/onboard...eway-service.ts	82%	81%	-1%
src/lib/inferen...board-probes.ts	92%	93%	+1%

---

_{Updated July 27, 2026 17:19 UTC}

[github-actions]

PR Review Advisor — No blocking findings reported

Advisor assessment: No blocking advisor findings reported
Next action: No advisor follow-up needed.
Findings: 0 blockers · 0 warnings · 0 suggestions

Model lanes

• GPT-5.6 Terra (primary): Completed · high confidence · 0 blockers · 0 warnings · 0 suggestions
• Nemotron 3 Ultra (second opinion): Completed · high confidence · 0 blockers · 0 warnings · 0 suggestions
• Model comparison: normalized findings match; normalized E2E selections differ; severity counts match.

Nemotron output stays in workflow artifacts and does not change the assessment above.

E2E guidance

Advisory only. E2E / PR Gate selects and runs jobs independently.

Recommended E2E: cloud-onboard, credential-sanitization, security-posture, inference-routing, network-policy

Workflow run details

This automated review informs maintainers. Warnings and suggestions do not require a response. A maintainer decides whether to merge.