[onmax]

Several assertions stopped before exercising Doctor: the OXC "parses scripts" test only queried a dependency capability, and an async-data probe only checked whether create() returned a visitor. This change removes the capability probe, consolidates repeated CLI/docs proof, and makes the existing six async-data smoke cases check actual diagnostics for ordinary and Unicode-escaped calls.

• tests/core/script-parse.test.ts: remove parses scripts with raw transfer where the runtime supports it. It never called Doctor and required host support even though Doctor supports a JSON fallback. The four remaining tests own AST fields, node stability, transfer parity, and parse errors through Doctor's parser.
• tests/vite/cli.test.ts: consolidate two repeated catalog invocations into the Vite-fixture catalog test. It now checks the specific Vite Rule's public package version, successful exit, and exclusion of Nuxt metadata, using the existing stdout helper.
• docs/rules/strict-only-rules.test.ts: merge the three exact command examples into the per-framework table that checks every Strict Preset member. Keep all exact example commands and assert each example remains in the Strict Preset so its assertion cannot silently stop running.
• tests/rule-packs/nuxt/nuxt-modern-rules-1.test.ts: reuse the six existing positive fixtures in test.each, asserting the exact diagnostic Rule and both file names. Remove 12 visitor-presence assertions; none dispatched a visitor, and the shared input could not trigger five of the six Rules. Retain the negative fast-path checks, which guard the performance regression from perf(nuxt,vue): skip files that cannot match in hot async-data and listener Rules #213, and existing alias/line-continuation cases.

Mutation control: temporarily returning null from getAsyncDataCall() left the original probe green, 1 passed. The repaired six cases all failed at the missing-diagnostic assertion under the same mutation. Production was restored byte for byte before normal validation.

Baseline is 593f9aed8787e75d0148e637cb409774dd4281ea on main. These measurements cover the four affected files:

Measure	Before	After
Passing tests	232	228
Sum of Vitest durations	26.35 s	11.99 s
Test LOC	3,805	3,771

Timings include startup and transforms, exclude the heavy queue, and are single runs on shared Slimbook with colder caches before. Repository-wide test LOC is 60,681 → 60,647. Production LOC remains 66,987; test support/fixture LOC remains 543. No production seams, fixtures, or dependencies became unused.

Final pnpm ready passed: 8,369 tests in 208 files, Vitest duration 89.11 s, plus formatting, build, lint, and type checking. Validated on Node 24.21.0 and pnpm 11.20.0.

Commands run:

HEAVY_WAIT=0 heavy pnpm install --frozen-lockfile

# Run once before and once after, for each file listed below.
HEAVY_WAIT=0 heavy --light pnpm test --run "$test_file" --maxWorkers=1 \
  --reporter=verbose --reporter=json --outputFile.json="$report"
# tests/core/script-parse.test.ts
# tests/vite/cli.test.ts
# docs/rules/strict-only-rules.test.ts
# tests/rule-packs/nuxt/nuxt-modern-rules-1.test.ts

# Deliberate mutation controls, restored afterward.
HEAVY_WAIT=0 heavy --light pnpm test --run tests/rule-packs/nuxt/nuxt-modern-rules-1.test.ts \
  --maxWorkers=1 --reporter=verbose \
  --testNamePattern='async data rules skip files that cannot call async data'
HEAVY_WAIT=0 heavy --light pnpm test --run tests/rule-packs/nuxt/nuxt-modern-rules-1.test.ts \
  --maxWorkers=1 --reporter=verbose \
  --testNamePattern='reports literal and escaped async-data calls'

HEAVY_WAIT=0 heavy --light pnpm exec vp fmt docs/rules/strict-only-rules.test.ts \
  tests/core/script-parse.test.ts tests/rule-packs/nuxt/nuxt-modern-rules-1.test.ts tests/vite/cli.test.ts
HEAVY_WAIT=0 heavy --light pnpm exec vp lint docs/rules/strict-only-rules.test.ts \
  tests/core/script-parse.test.ts tests/rule-packs/nuxt/nuxt-modern-rules-1.test.ts tests/vite/cli.test.ts
HEAVY_WAIT=0 heavy pnpm ready --run --maxWorkers=1 --no-file-parallelism \
  --reporter=verbose --reporter=json --outputFile.json=artifacts/test-audit/full-after.json
git diff --check

pnpm ready formats, builds, lints with typeAware and typeCheck enabled in vite.config.ts, and runs the full suite. OpenClaw-specific deadcode/check-changed commands do not exist in this repository.

Kept the Nuxt fixture integration tests: setup() is correctly registered at describe scope. Also retained the security override checks, parser/platform contracts, typed public APIs, Vue runtime examples, and parse-count/scaling guards. Their distinct contracts or historical regressions did not justify deletion.

🤖 Generated with Claude Code

[chatgpt-codex-connector]

Codex Review Summary

This comment shows the latest Codex review activity on this pull request.

Review	Status	Commit	Review trigger
📝 Code Review	✅ Completed 2026-11-01T10:22:28.233891Z	a278093	PR opened

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.

[vite-hub]

Session	Status	Started	Duration
Pending	Completed	Not started	—
View session	Waiting	2026-11-01T10:27:28.227Z	11m 6s

Latest result

Pull request merged.

Final answers

Session

Pull request merged.

Session

Waiting for GitHub merge request ba409aad-771d-4c3d-9bc4-a5ea14d03123.

View session

Ready: expected HEAD verified, all supplied checks passed, and no actionable review findings remain. Focused pnpm tests passed all 228 tests across the four changed files; focused vp lint and git diff --check passed. Validation used --config.verify-deps-before-run=false to avoid pnpm’s automatic-reinstall error. The worktree is clean; no repair commit was needed.

Session

Waiting for review checks before the next pass.

Session

Waiting for review checks before the next pass.

[pullfrog]

✅ No new issues found.

Reviewed changes across the four test suites, including whether the consolidated assertions preserve the independent parser, catalog, documentation, and Nuxt async-data contracts.

• Async-data diagnostics — replaces visitor-presence probes with end-to-end Diagnostic assertions for six Rules, covering both literal and Unicode-escaped composable calls.
• Vite catalog coverage — consolidates duplicate CLI invocations while retaining exit status, Vite Rule presence, public package version, and Nuxt metadata exclusion.
• Strict command coverage — combines exact command examples with complete Strict Preset membership checks for TypeScript, Vite, and Vue.
• Parser capability probe — removes a host-level raw-transfer assertion while retaining Doctor-level AST, stability, parity, and parse-error coverage.

Focused verification passed: 228 tests across all four changed files.

  ^{｜ View workflow run ｜ Using gpt-5.6-sol ｜ 𝕏}

[pkg-pr-new]

Open in StackBlitz

pnpm dlx https://pkg.pr.new/vite-doctor@309

commit: a278093