Skip to content

chore(deps): group low-risk frontend bumps (ai, eslint, vite, types) - #229

Merged
westonplatter merged 1 commit into
mainfrom
chore/deps-bun-low
Sep 3, 2026
Merged

chore(deps): group low-risk frontend bumps (ai, eslint, vite, types)#229
westonplatter merged 1 commit into
mainfrom
chore/deps-bun-low

Conversation

@westonplatter

Copy link
Copy Markdown
Owner

Summary

Batches the eight open low-tier bun dependabot PRs into one reviewable change, and picks up a high-severity transitive advisory along the way. bun add writes exact pins and discards the caret, so the ranges were restored afterwards — package.json moves only the eight version floors, nothing changed dep/devDep sections.

Package From To Kind Supersedes
ai 7.0.62 7.0.68 runtime patch #208
@ai-sdk/react 4.0.65 4.0.71 runtime patch #214
globals 17.8.0 17.10.0 dev minor #211
eslint 10.8.0 10.8.1 dev patch #212
@vitejs/plugin-react 6.0.4 6.0.5 dev patch #215
eslint-plugin-react-refresh 0.5.3 0.5.4 dev patch #216
@types/node 26.1.2 26.2.0 type stubs minor #217
vite 8.2.0 8.2.1 dev patch #218

Features

N/A

Refactoring

N/A

Fixes

  • Drops GHSA-2v37-7h3g-55p8 (nanoid, high — custom generators loop indefinitely when size is zero). No bot PR could reach it: nanoid has no direct entry and sits under vite › postcss. vite 8.2.1 pulls postcss 8.5.26, which requires nanoid >=3.3.18, so the advisory resolves without an overrides entry. Audit goes 5 → 4.

Verification

Measured with compare_baseline.py --adapter bun, which diffs a detached worktree at origin/main against this tree:

metric origin/main this branch verdict
build (tsc -b + vite build) 0 errors 0 errors same
lint (eslint .) 0 errors 0 errors same
audit (bun audit) 5 vulnerabilities 4 better (−1)

Three of the eight (eslint, eslint-plugin-react-refresh, vite) are in the adapter's baseline_tools — they produce the counts above, so an unchanged number is not by itself evidence. The findings were diffed directly:

  • bun run lint output is byte-identical before and after.
  • bun run build output differs only in the vite v8.2.0v8.2.1 banner; the emitted asset list and hashes are unchanged.

Also run locally, matching the CI frontend job: bun run typecheck (clean) and bun test (71 pass, 0 fail).

Lock blast radius: 13 entries moved for 8 requested packages. The 5 extras are each parented by one of the eight, not a graph re-resolve — @ai-sdk/gateway 4.0.54 via ai, @ai-sdk/mcp 2.0.33 via @ai-sdk/react, jiti 2.7.0 via eslint/vite, postcss 8.5.26 and nanoid 3.3.18 via vite.

Cooldown: all thirteen versions — the eight direct bumps and the five transitive extras that no bot PR ever proposed — are past the repo's 14-day window (oldest 15d, ai and @ai-sdk/*; bunfig.toml minimumReleaseAge enforces this mechanically on the direct installs).

Documentation

None needed — no behavior change.

Additional notes

🤖 Generated with Claude Code

https://claude.ai/code/session_011wzZeis7evtxWQtZaQWis6


Generated by Claude Code

Batches the eight low-tier bun dependabot PRs into one change. Caret
ranges are restored after `bun add` writes exact pins, so only the
version floors move in package.json.

- ai 7.0.62 -> 7.0.68 (supersedes #208)
- @ai-sdk/react 4.0.65 -> 4.0.71 (supersedes #214)
- globals 17.8.0 -> 17.10.0 (supersedes #211)
- eslint 10.8.0 -> 10.8.1 (supersedes #212)
- @vitejs/plugin-react 6.0.4 -> 6.0.5 (supersedes #215)
- eslint-plugin-react-refresh 0.5.3 -> 0.5.4 (supersedes #216)
- @types/node 26.1.2 -> 26.2.0 (supersedes #217)
- vite 8.2.0 -> 8.2.1 (supersedes #218)

Verified against the origin/main baseline: build 0 errors (same), lint 0
errors (same), audit 5 -> 4 vulnerabilities. eslint,
eslint-plugin-react-refresh and vite produce those counts, so the
findings were diffed rather than the counts alone: eslint output is
byte-identical, and the build output differs only in the vite version
banner.

The dropped advisory is GHSA-2v37-7h3g-55p8 (nanoid, high) - vite 8.2.1
pulls postcss 8.5.26, which requires nanoid >=3.3.18. Thirteen lock
entries moved: the eight above plus five internals each parented by one
of them (@ai-sdk/gateway via ai, @ai-sdk/mcp via @ai-sdk/react, jiti via
eslint and vite, postcss and nanoid via vite). All five are past the
14-day cooldown, as are the eight direct bumps.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_011wzZeis7evtxWQtZaQWis6
@coderabbitai

coderabbitai Bot commented Sep 3, 2026

Copy link
Copy Markdown
Contributor

Important

Review skipped

Auto reviews are disabled on this repository. Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Team

Run ID: 9be90a75-6f5e-474d-bc52-ba99a0b06f01

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

  • 🔍 Trigger review

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

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

@westonplatter
westonplatter merged commit bc86752 into main Sep 3, 2026
3 checks passed
@westonplatter
westonplatter deleted the chore/deps-bun-low branch September 3, 2026 04:15
westonplatter pushed a commit that referenced this pull request Sep 3, 2026
Resolves the conflict #229 created: both branches moved adjacent lines in
frontend/package.json and re-resolved frontend/bun.lock from the same
base.

Resolution keeps both sides - typescript-eslint ^8.67.0 from this branch,
vite ^8.2.1 and the other seven bumps from main. The lock was regenerated
with `bun install` rather than hand-merged; it now differs from main by
exactly the 11 typescript-eslint entries this PR is for.

Re-verified against the new origin/main baseline (which now carries #229):
build 0 errors (same), lint 0 errors (same), audit 4 vulnerabilities
(same - the nanoid advisory #229 dropped stays dropped). eslint output is
byte-identical, typecheck clean, bun test 71 pass.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_011wzZeis7evtxWQtZaQWis6
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants