Repository navigation
build(deps): bump hono from 4.13.4 to 4.13.11 in /apps/api - #59
dependabot[bot] wants to merge 6 commits into
Conversation
|
CI is red because this PR only touched @dependabot recreate |
301383b to
75f097e
Compare
|
CI is red across almost every job (Test, Build, Lint, Secret Scan, etc.), not just E2E Smoke — root cause is the same for all of them: this PR only bumped @dependabot recreate |
Bumps [hono](https://github.com/honojs/hono) from 4.13.4 to 4.13.11. - [Release notes](https://github.com/honojs/hono/releases) - [Commits](honojs/hono@v4.13.4...v4.13.11) --- updated-dependencies: - dependency-name: hono dependency-version: 4.13.11 dependency-type: direct:production ... Signed-off-by: dependabot[bot] <support@github.com>
75f097e to
2f7b9e7
Compare
dependabot's lockfile update doesn't resolve this repo's pnpm workspace/catalog layout, so #59 left pnpm-lock.yaml stale and broke every CI job at the frozen-lockfile install step (two @dependabot recreate attempts reproduced the same stale lockfile). Regenerated with pnpm install --no-frozen-lockfile using the pinned pnpm 9.15.9. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01Y8g7sNrGUynUGGRNRtTuPc
|
The `@dependabot recreate` above didn't fix it — the regenerated commit (2f7b9e7) still only touched `apps/api/package.json`, not `pnpm-lock.yaml`. This looks like a repo-wide gap in dependabot's pnpm workspace/catalog resolution here, not something retrying will fix (same signature as #52's lockfile issue, but #52's own commit happened to include the lockfile — this one and #56 don't). Pushed `381b2591b` directly to this branch with a regenerated `pnpm-lock.yaml` (via `pnpm install --no-frozen-lockfile`, pinned pnpm 9.15.9, verified `--frozen-lockfile` now passes locally). CI should go green on the new commit; the hono/valibot peer bump is otherwise untouched. 🤖 Generated with Claude Code |
…ity gate
hono 4.13.11 tightens TS2742 (inferred type cannot be named without a
non-portable reference) on two handlers whose return type depended on
hono's internal, unexported types: deleteNodeRoute (annotated
Promise<Response>) and jsonValidator (a new JsonValidatorMiddleware<T>
alias built from hono's public MiddlewareHandler/Input types, since
@hono/valibot-validator's own inferred return type routes through an
unexported context type). No behavioral change; `c.req.valid('json')`
narrowing for jsonValidator consumers is preserved.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Y8g7sNrGUynUGGRNRtTuPc
|
Lockfile was fixed, but Type Check/Test/Build still failed for a separate reason: hono 4.13.11 tightens a TS2742 check ("inferred type cannot be named without a non-portable reference") on two handlers whose signatures depended on hono's internal, unexported types — `deleteNodeRoute` (`src/routes/nodes/delete.ts`) and `jsonValidator` (`src/schemas/_validator.ts`, via `@hono/valibot-validator`'s own inferred return type). Pushed `c0cfc832c` adding explicit, portable return-type annotations for both (no behavior change — `c.req.valid('json')` narrowing for `jsonValidator` callers is preserved via a new `JsonValidatorMiddleware` alias built from hono's public types). Verified locally: `tsc --noEmit` and `pnpm build` both pass clean in `apps/api`. Secret Scan's remaining failure ("A reviewed Gitleaks baseline entry has expired") is unrelated and repo-wide — tracked separately on #57. CodSpeed's failure is the known pre-existing 401 (#46), also unrelated to this bump. 🤖 Generated with Claude Code |
…rn/apps/api/hono-4.13.11 # Conflicts: # pnpm-lock.yaml
|
Root cause of the stuck checks: this PR had drifted far enough behind `main` (mastra/core, ai, workers-ai-provider bumps landed in the meantime) that it became unmergeable, and GitHub doesn't run `pull_request`-triggered CI on a branch it can't form a merge ref for — that's why no check ran on the previous commit for two days, not an Actions outage. Pushed `e832091`, merging `main` in and regenerating the lockfile. CI is running now on the merged commit. 🤖 Generated with Claude Code |
…d it before regenerating) The prior merge commit's git add ran before pnpm install regenerated the lockfile, so it committed the pre-merge lockfile (still specifying hono@4.13.4) despite package.json already specifying 4.13.11. CI failed with the same ERR_PNPM_OUTDATED_LOCKFILE this PR already fixed once. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01Y8g7sNrGUynUGGRNRtTuPc
|
My merge commit (e832091) had a bug: `git add pnpm-lock.yaml` ran before `pnpm install --no-frozen-lockfile` regenerated it, so it committed the pre-merge lockfile (still `hono@4.13.4`) despite `package.json` already saying `4.13.11` — same `ERR_PNPM_OUTDATED_LOCKFILE` as before, caused by me this time, not dependabot. Pushed `ce1aacd` with the lockfile actually regenerated and staged in the right order. Verified locally: `pnpm install --frozen-lockfile`, `tsc --noEmit`, and `pnpm build` all pass. CI is running now. 🤖 Generated with Claude Code |
…iddleware
@typescript-eslint/no-explicit-any flagged the any used for the Handler's
Bindings/Variables generic. hono's own Env type ({Bindings?, Variables?})
is the correct unconstrained default and keeps the lint rule satisfied
without widening to any.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Y8g7sNrGUynUGGRNRtTuPc
|
Lint was still red (`@typescript-eslint/no-explicit-any` on the `any` I used for `JsonValidatorMiddleware`'s Env generic). Pushed `e15bb4f` replacing it with hono's own `Env` type (`{ Bindings?; Variables? }`), which is the correct unconstrained default. Verified locally: `tsc --noEmit`, `eslint`, and `pnpm build` all pass clean in `apps/api`. The `Test` failure on the previous run (`automation-flows.test.tsx`, a 5s timeout on 1 of 4432 tests) is an unrelated CI-load flake, not caused by this bump — it's in web UI code this PR never touches. `Preflight Evidence` is the known always-red-on-bot-PRs gate. 🤖 Generated with Claude Code |
|
CI is green on `e15bb4f`: Lint, Type Check, Test, Build, Secret Scan, and all Durable Object Workers shards all pass. The two remaining red checks are pre-existing and unrelated to this bump — Preflight Evidence (known always-red on bot/dependabot PRs) and CodSpeed's benchmark run (known 401 since 2026-07-11, #46). This PR should be mergeable now. 🤖 Generated with Claude Code |
Bumps hono from 4.13.4 to 4.13.11.
Release notes
Sourced from hono's releases.
... (truncated)
Commits
37ce0694.13.111e1207ctest(serve-static): fix the test (#5479)8b05c77Merge commit from fork18331a9chore(adapters): release@hono/netlify@1.0.16cd583bchore(adapters): release@hono/deno@1.0.2271dd0bci: write the tag version into deno.json before publishing to JSR (#5469)ae93026chore(adapters): release@hono/deno@1.0.172f0d8dfeat(adapters/netlify): publish to JSR (#5468)f3d5e884.13.10645ddd6test: move adapter runtime tests intoadapters/*(#5466)