test(sandbox): share the snapshot restore test fixture

[laitingsheng]

Summary

The snapshot suite carried its own copy of the mock prologue that snapshot-restore-test-fixture.ts already owns and that eight sibling suites already share. The suite now imports that fixture instead, so one module owns the shared mock registrations and their reset.

Related Issue

Resolves #8349

Changes

• snapshot.test.ts drops the duplicated prologue: two types, three openshell response helpers, two vi.hoisted blocks, 22 mock constants, 16 vi.mock registrations, a 33-line beforeEach reset body, and the afterEach. It now uses import * as f from "./snapshot-restore-test-fixture" with f.resetSnapshotRestoreMocks() and f.cleanupSnapshotRestoreMocks(), matching the eight suites that already consume the fixture.
• Scenario inputs, test titles, and assertions are unchanged. Mock references in the body are reached through the f. namespace.
• Suite-local helpers stay local: dcodeSandboxEntry, mockDcodeProbe, mockDcodeProbeResult, capturedDcodeProbeScript, and runProbeScriptWithProcesses.
• snapshot-restore-test-fixture.ts delegates resolveAgentBaselinePolicyMock to resolveTestAgentBaselinePolicy, replacing a byte-identical private copy. test/support/snapshot-policy-test-fixture.ts stays the owner of baseline-policy selection. This matters because a fixture vi.mock registration runs after the importing suite's hoisted registrations and wins for the same path, so the suite cannot keep a competing ../../policy mock.
• snapshot.test.ts goes from 1476 to 1220 lines. Net 269 fewer lines. No production source changed and no production export added.
• resolveTestAgentBaselinePolicy in test/support/snapshot-policy-test-fixture.ts now matches the production resolveAgentBaselinePolicy contract: it accepts string | null | undefined, falls back to openclaw for a missing agent the same way production does, and declares the nullable return so a suite can model the unreadable-baseline path.
• The restore lifecycle test asserts that harden, delete, and cleanup-shields were all recorded before it compares their positions, so a missing event can no longer satisfy the ordering through indexOf returning -1.

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: this is a test-only consolidation; every case, title, and assertion is preserved, and the 125 tests across the 10 suites that share the fixture pass unchanged.
• Tests not applicable — justification:
• Docs updated for user-facing behavior changes
• Docs not applicable — justification: no public API, CLI, configuration, default, or error message changes.
• 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: nine-category security review PASS with no findings; all changed files are tests or test support, with no production behavior changes. Evidence: test(sandbox): share the snapshot restore test fixture #8447 (comment)
• 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: The full PR and current follow-up change only tests and test support. No CLI behavior, configuration, policy schema, default, documentation source, or other user-facing contract changes. The independent reviewer inspected the production resolver, full PR diff, local follow-up, docs skip list, and documentation contributor guidance.
• Agent: Codex documentation writer

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 validate:pr passed after refreshing origin/main 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: npx vitest run --project cli src/lib/actions/sandbox/snapshot.test.ts src/lib/actions/sandbox/snapshot-restore-*.test.ts src/lib/actions/sandbox/snapshot-managed-*.test.ts — 10 files, 125 tests passed; after the merge and the review fixes, npx vitest run --project cli src/lib/actions/sandbox — 232 files, 2524 of 2526 tests passed, with the two remaining tests timing out at the 5s default under local parallel load and passing when their files run on their own; npm run typecheck:cli clean; npm run validate:pr clean; npm run test:projects:check reports exact membership; npx biome check clean on the changed files.
• 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: Tinson Lai tinsonl@nvidia.com

Summary by CodeRabbit

• Tests
  • Expanded snapshot coverage for creation, restoration, error handling, policy reconciliation, backup failures, timer and shield ordering, and non-live sandbox scenarios.
  • Standardized snapshot tests on shared test fixtures and response helpers for more consistent validation.
  • Added coverage confirming baseline-policy handling for missing, null, and valid agent values.
• Refactor
  • Consolidated baseline-policy test behavior through shared testing utilities.
  • Simplified snapshot test setup by reusing common test infrastructure.

[coderabbitai]

📝 Walkthrough

Walkthrough

Snapshot tests now use shared fixtures and helpers instead of local definitions. The suite covers DCode probes, snapshot restore lifecycle behavior, clone observability, and OTLP policy reconciliation. The baseline-policy resolver and mock now support nullish agents.

Changes

Snapshot fixture consolidation

Layer / File(s)	Summary
Baseline policy fixture reuse src/lib/actions/sandbox/snapshot-restore-test-fixture.ts, test/support/snapshot-policy-test-fixture.ts, src/lib/actions/sandbox/snapshot-restore-baseline-exclusions.test.ts	The baseline-policy resolver handles nullish agents and has explicit return typing. The resolver contract is tested. The baseline-policy mock and reset behavior use the shared resolver.
Shared snapshot test infrastructure src/lib/actions/sandbox/snapshot.test.ts	Snapshot tests use shared fixture data, mocks, types, and response helpers for shield, snapshot, listing, timer, and DCode probe scenarios.
Restore lifecycle coverage src/lib/actions/sandbox/snapshot.test.ts	Tests cover serialization, permission repair, destination handling, gateway conflicts, and cloning with shared lifecycle and sandbox fixtures.
Restore policy coverage src/lib/actions/sandbox/snapshot.test.ts	Tests cover clone observability, OTLP policy reconciliation, custom-policy ownership, replay failures, non-live sandboxes, and backup errors.

Estimated code review effort: 3 (Moderate) | ~30 minutes

Possibly related issues

• Make the snapshot suite use the existing restore test fixture #8349: The PR refactors snapshot.test.ts to use the existing restore test fixture and shared snapshot-policy fixture.

Possibly related PRs

• NVIDIA/NemoClaw#8430: Both PRs refactor shared sandbox test fixtures and centralize reusable utilities.

Suggested reviewers: cv, prekshivyas, cjagwani

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Docstring Coverage	✅ Passed	Docstring coverage is 100.00% which is sufficient. The required threshold is 80.00%.
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.
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title clearly and concisely describes the main change: sharing the snapshot restore test fixture.

✨ 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/snapshot-suite-restore-fixture

---

_{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 8d4c647 in the test/snapshot-suite-... branch remains at 96%, unchanged from commit e323de1 in the main branch.

TypeScript / code-coverage/cli

The overall coverage in commit 8d4c647 in the test/snapshot-suite-... branch remains at 81%, unchanged from commit e323de1 in the main branch.

Show a code coverage summary of the most impacted files.

File	main e323de1	test/snapshot-suite-... 8d4c647	+/-
src/lib/shields/index.ts	83%	82%	-1%
src/lib/actions...air-approval.ts	90%	89%	-1%
src/lib/inferen...ompatibility.ts	94%	94%	0%
src/lib/actions...test-fixture.ts	91%	91%	0%
src/lib/messagi...an-authority.ts	100%	100%	0%
src/lib/sandbox...rce-identity.ts	87%	87%	0%
src/lib/domain/...dbox/destroy.ts	84%	89%	+5%
src/lib/adapter...regular-file.ts	93%	98%	+5%
src/lib/inferen...a/model-size.ts	83%	96%	+13%

---

_{Updated August 06, 2026 14:23 UTC}

[coderabbitai]

Actionable comments posted: 2

🤖 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 `@src/lib/actions/sandbox/snapshot-restore-test-fixture.ts`:
- Line 160: Update resolveTestAgentBaselinePolicy and its
resolveAgentBaselinePolicyMock wrapper to match the production resolver
contract: accept string, null, or undefined, and return null when no baseline is
provided while preserving the object result for valid strings.

In `@src/lib/actions/sandbox/snapshot.test.ts`:
- Around line 648-653: Update the lifecycle-order assertions in the snapshot
test to first verify that “harden”, “delete”, and “cleanup-shields” are present
in f.lifecycleMock.events, then compare their indices. Keep the existing
ordering requirements while preventing missing events from satisfying the test
through indexOf returning -1.

🪄 Autofix

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: 2b5eb63e-d269-4772-8b27-156863ab7b85

📥 Commits

Reviewing files that changed from the base of the PR and between e40ca39 and ae3d629.

📒 Files selected for processing (2)

• src/lib/actions/sandbox/snapshot-restore-test-fixture.ts
• src/lib/actions/sandbox/snapshot.test.ts

[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 terminology decisions differ; normalized E2E selections match; severity counts match.

Second-opinion terminology and E2E selections are advisory. Live E2E does not run automatically for pull requests.

1 semantic terminology decision

Terminology decisions are advisory. They affect the assessment only when a separate finding identifies concrete semantic impact.

• established — OpenClaw baseline at src/lib/actions/sandbox/snapshot-restore-baseline-exclusions.test.ts:12: Keep “OpenClaw baseline”; it consistently identifies the OpenClaw-specific case of the established baseline policy term.

E2E guidance

Advisory only. A maintainer can dispatch the default E2E suite against this exact revision.

Recommended E2E: onboard-repair, onboard-resume, rebuild-openclaw, state-backup-restore

1 optional E2E recommendation

• snapshot-commands

Workflow run details

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

[coderabbitai]

Actionable comments posted: 1

🤖 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 `@src/lib/actions/sandbox/snapshot-restore-baseline-exclusions.test.ts`:
- Around line 12-20: Update the test around “preserves the nullable production
baseline resolver contract” to exercise the real resolver through the
snapshot-restore public entrypoint, passing nullish and valid agent values and
asserting the observable results. Do not invoke resolveAgentBaselinePolicyMock
directly; if that boundary cannot expose the contract, rename the test to
explicitly describe fixture-helper behavior instead.

🪄 Autofix

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: 56b61818-d09d-4d71-be2d-a3788605800f

📥 Commits

Reviewing files that changed from the base of the PR and between e5455b1 and 2526d9d.

📒 Files selected for processing (2)

• src/lib/actions/sandbox/snapshot-restore-baseline-exclusions.test.ts
• test/support/snapshot-policy-test-fixture.ts

[apurvvkumaria]

CI classification update: every required check is green for the current branch revision.

The second-opinion advisor lane ended in receipt validation after producing two suggestions. Both suggestions describe work already present in this revision: the snapshot fixture consolidation and the nullable fallback contract aligned with production. The primary advisor and publisher completed, the actionable warning was addressed with regression coverage, and its review thread is resolved. There is no remaining actionable advisor finding, so I am not rerunning the unchanged second-opinion lane.

The focused snapshot suite, type checks, repository gates, DCO, commit verification, documentation review, and security review all pass. The remaining merge gate is independent maintainer approval.

[apurvvkumaria]

Automated-review follow-up for the current branch revision (8d4c647): the shared snapshot baseline fixture now matches production fallback behavior for absent agent metadata, selecting the OpenClaw baseline for both null and undefined inputs. The regression test is explicitly scoped to the fixture helper and covers null, undefined, and explicit OpenClaw values. This addresses the primary advisor warning and the new CodeRabbit thread; the thread has a direct response and is resolved.

Validation passed: affected snapshot suites 47/47, CLI typecheck, test-title policy, source-shape budget, test-size budget, normal commit hooks, and pre-push checks.

Independent documentation review: PASS, no docs needed. The full PR and follow-up change only tests and test support; no CLI behavior, configuration, policy schema, default, or docs source changes.

Nine-category security review: PASS with no findings. There are no production runtime, secret, input-processing, authorization, dependency, logging, cryptography, or security-configuration changes. The follow-up improves security-test fidelity by removing a fixture fallback that diverged from production. Fresh CI and automated review are running.