chore(deps): clean npm audit and bump puppeteer to 25.10.0 - #19
Merged
Conversation
Resolves all 4 advisories (3 high, 1 moderate) and brings puppeteer to the current release. The audit findings were entirely transitive and none came from puppeteer: fast-uri 3.1.2 -> 3.1.7 (openapi-mock-validator > ajv) js-yaml 4.3.0 -> 4.3.2 (openapi-mock-validator > json-schema-ref-parser) nanoid 3.3.16 -> 3.3.18 (vitest > vite > postcss, dev only) postcss 8.5.20 -> 8.5.28 (vitest > vite, dev only) All four resolved in-range via `npm audit fix`, no --force and no direct dependency change. puppeteer 25.3.0 -> 25.10.0 is audit-neutral and was validated separately against the recording feature, since that is the part that leans on undocumented screencast internals: - ScreenRecorder's frame pipeline is unchanged, `bufferCount(2, 1)` included, so the holdFinalFrame workaround still holds. - `ffmpegPath` and the `setpts` speed filter are unchanged. - `path`/`overwrite` moved out of ScreenRecorder and into Page.screencast, which now writes through a stream instead of ffmpeg's -y flag. Verified by recording over a planted oversized file: fully truncated, no tail. - `page.screencast()` is now @deprecated in favour of `Page.record()`, but JSDoc-only. No runtime warning, so CLI output stays quiet. Verified end to end with real recordings, not just the mocked suite: identical codec/resolution/fps, and 128 frames on both versions for the same 15s run. Frame counts on short runs vary run to run on both (25.3.0: 28-39, 25.10.0: 28-33), so there is no capture regression. vitest and @vitest/coverage-v8 took their in-range 4.1.10 -> 4.1.11. vitest 5.0.0 is left alone: a dev-only major with no audit benefit. Lockfile regenerated with `npm run lock:linux` and `npm ci` verified in a node:24 container. The @emnapi/* entries are gone rather than stale this time: vite 8.2.2 pulls a rolldown that no longer ships the wasm32-wasi binding those were dependencies of. 439 tests pass, 0 vulnerabilities.
TWD Contract Validation
23 passed · 41 failed · 3 warnings · 1 skipped Failed validations./contracts/users-3.0.json
./contracts/posts-3.1.json
./contracts/products-3.0.json
./contracts/events-3.1.json
|
This was referenced Sep 6, 2026
Merged
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Clears all 4 npm audit advisories (3 high, 1 moderate) and brings puppeteer to the current release.
The two are independent decisions, and it's worth separating them when reviewing: none of the advisories came from puppeteer.
Clean audit — lockfile only, zero risk
openapi-mock-validator→ ajvopenapi-mock-validator→ json-schema-ref-parservitest→ vite → postcss (dev only)vitest→ vite (dev only)All four resolved in-range via
npm audit fix— no--force, no direct dependency change.package.jsonwas untouched by this half.puppeteer 25.3.0 → 25.10.0 — audit-neutral, validated separately
This is where the actual risk sits, since the recording feature leans on undocumented screencast internals. Reviewed by diffing puppeteer's
ScreenRecorder.tsandPage.tsbetween the two versions rather than trusting release notes:bufferCount(2, 1)is byte-identical, so theholdFinalFrameworkaround still holds.ffmpegPathand thesetptsspeed filter are also unchanged.path/overwritegenuinely moved out ofScreenRecorderintoPage.screencast(), which now writes through a stream instead of ffmpeg's-yflag. Verified by recording over a planted oversized 3 MB file: fully truncated, no leftover tail.page.screencast()is now@deprecatedin favour ofPage.record(). JSDoc-only, no runtime warning, so CLI output stays quiet.Recording verified for real, not just mocked
The unit suite mocks
page.screencast(), so it proves nothing here. Ran actual recordings on both versions:--record-speed 2still halves duration; extracted frames are valid, sidebar hidden, html margin reset intact.Also included
vitestand@vitest/coverage-v8took their in-range 4.1.10 → 4.1.11.Deliberately not included: vitest 5.0.0 — a dev-only major with no audit benefit. Better as its own branch where a break is obviously the upgrade's fault rather than mixed into a security fix.
Lockfile note
Regenerated with
npm run lock:linux, andnpm civerified in anode:24container (exit 0).The
@emnapi/*entries disappeared, but not as the usual macOS staleness — vite 8.2.2 pulls a rolldown that no longer ships thewasm32-wasibinding those were dependencies of. They left the tree legitimately.Verification
npm audit→ 0 vulnerabilitiesnpm cion linux/node:24 → exit 0Follow-up (not in this PR)
Page.record()would remove the ffmpeg dependency entirely — the pre-flight probe,assertFfmpegAvailable(), the install-help message,record.ffmpegPath. But it has nospeedoption, so--record-speedwould be lost;pacearguably covers that need better anyway. Worth its own spike.🤖 Generated with Claude Code