chore(deps): patch hono, and drop two overrides that pinned nothing - #380
Conversation
Audited all 13 root overrides against what the lockfile actually resolves, after finding that the Desktop runtime had never had any of them (#376). Three results. `hono` was pinned `^4.13.0` and resolving to 4.13.1, which is inside the advisory range for both the `toSSG()` write and the query-parser issue (fixed in 4.13.5). Pinned `^4.13.5`; it resolves to 4.13.7. Three lines of lockfile. `basic-ftp` and `jayson > uuid` pin nothing — none of `basic-ftp`, `jayson` or `uuid` is in the tree at all. They are leftovers from dependencies that have since gone, and an override that matches nothing is worse than no override: it reads as protection. Removed, with no change to any resolution. The other nine are doing their job: axios 1.19.0, undici 8.10.0, ws 8.21.3, postcss 8.5.26, esbuild 0.28.1, brace-expansion 5.0.9, fast-uri 3.1.5, tar 7.5.22, ip-address 10.4.0 — one copy each. NOT bumped: vitest `^4.1.3` (resolves 4.1.10) against the @vitest/mocker path-traversal advisory, fixed in 4.1.11. Every route to it — `vitest@^4.1.11`, `vitest@4.1.11` exact — crashes Arborist in `#loadPeerSet` with "Cannot read properties of null (reading 'edgesOut')", the same failure as #373, and retrying does not clear it; unmodified main resolves fine. vitest declares exact-version peers on a dozen optional `@vitest/*` packages, which is the likely trigger. It is development scope and not shipped, so it is not worth a hand-edited lockfile on a package that publishes real-money code. Left for when npm or vitest moves. Verified by clean `npm ci` from this lockfile in a scratch tree: hono 4.13.7, tsc clean, 1088 tests pass. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_015mUab3xrLHNpYqVHJLgJHL
|
No actionable comments were generated in the recent review. 🎉 ℹ️ Recent review info⚙️ Run configurationConfiguration used: Path: .coderabbit.yaml Review profile: CHILL Plan: Advanced Run ID: ⛔ Files ignored due to path filters (1)
📒 Files selected for processing (1)
Included review availability: Your plan provides up to 8 included reviews per hour; 3 remain after this review. 📝 WalkthroughWalkthroughThe package override configuration removes the ChangesDependency override updates
Estimated code review effort: 1 (Trivial) | ~2 minutes Suggested reviewers: Merge Risk: ⚪ Minimal · up to This updates the Hono security patch and removes overrides for absent packages. Dependency resolution is consistent with the new Hono floor, with no current merge-blocking risk identified. 🚥 Pre-merge checks | ✅ 5✅ Passed checks (5 passed)
✨ Finishing Touches🧪 Generate unit tests (beta)
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. Comment |
Follow-on from #376. Having found that the Desktop runtime had never had any of the root security pins, the obvious next question was whether the root ones are doing what they claim. Audited all 13 against what the lockfile actually resolves.
hono was inside the advisory range
Pinned
^4.13.0, resolving to 4.13.1. Both open hono advisories (thetoSSG()write and the query parser reading past the URL fragment) are fixed in 4.13.5. Pinned^4.13.5; resolves to 4.13.7. Three lines of lockfile.Two overrides pin nothing
basic-ftpandjayson > uuidmatch no package in the tree — none ofbasic-ftp,jaysonoruuidis installed at all. Leftovers from dependencies that have since gone. An override matching nothing is worse than no override, because it reads as protection: #358's body cites "the 13 entries already there are security pins … all still enforced", and two of them enforce nothing. Removed; no resolution changes.The other nine are fine
One copy each, all above their advisories:
axios1.19.0,undici8.10.0,ws8.21.3,postcss8.5.26,esbuild0.28.1,brace-expansion5.0.9,fast-uri3.1.5,tar7.5.22,ip-address10.4.0.Worth knowing while reading the security tab: the four
fast-uri"high" alerts are phantom. They cite ranges>= 2.3.1, < 2.4.5, and the tree contains exactly onefast-uri, at 3.1.5 (fast-uri/-/fast-uri-3.1.5.tgz, the only tarball in the lockfile). They should clear on the next dependency-graph pass.What is deliberately not fixed
vitest^4.1.3resolves to 4.1.10, and the@vitest/mockerpath-traversal advisory is fixed in 4.1.11. Every route to it —vitest@^4.1.11, exactvitest@4.1.11— crashes Arborist:Same failure as #373, and retrying does not clear it; unmodified
mainresolves fine, so the bump is the trigger. vitest declares exact-version peers on a dozen optional@vitest/*packages, which is the likely cause. It is development scope and not shipped, and hand-editing a lockfile on a package that publishes real-money code to chase a test-runner advisory is a bad trade. Left for when npm or vitest moves.Verification
Clean
npm cifrom this lockfile into a scratch tree:hono4.13.7,tsc --noEmitclean, 1088 tests pass.🤖 Generated with Claude Code
https://claude.ai/code/session_015mUab3xrLHNpYqVHJLgJHL
Summary by CodeRabbit