Skip to content

chore(deps): update all non-major dependencies - #58

Merged
RedStar071 merged 1 commit into
mainfrom
renovate/all-minor-patch
Nov 16, 2026
Merged

RedStar071 merged 1 commit into
mainfrom
renovate/all-minor-patch

Conversation

@renovate

@renovate renovate Bot commented Nov 15, 2026 •

Copy link
Copy Markdown
Contributor

Note

This PR body was truncated due to platform limits.

This PR contains the following updates:

Package Type Update Change Pending Age Confidence
actions/setup-node action minor v7.0.0 → v7.1.0 age confidence
aube tools minor 2.9.0 → 2.10.0 age confidence
docker/login-action action minor v4.7.0 → v4.8.0 age confidence
oxfmt (source) devDependencies minor ^0.76.0 → ^0.78.0 age confidence
pullfrog/pullfrog action patch v0.1.90 → v0.1.129 v0.1.130 age confidence

Release Notes

actions/setup-node

v7.1.0

Compare Source

What's Changed
Enhancements:
Bug fixes:
Documentation updates:
Dependency updates:
New Contributors

Full Changelog: actions/setup-node@v7.0.0...v7.1.0

aubepkg/aube (aube)

v2.10.0: : npm shim workspace flags, workspace-wide audit, and lockfile compatibility fixes

Compare Source

The npm shim now handles npm's --workspace/--workspaces, --audit, and --fund flags. aube audit now checks every workspace package, and this release fixes several lockfile, resolver, and registry bugs.

Added
  • --no-include-workspace-root (#​1917, @​jdx): this hidden global flag leaves the workspace root out of recursive selections unless a selector names it directly. It overrides includeWorkspaceRoot: true and --include-workspace-root.
    aube -r --no-include-workspace-root run build
Fixed
  • npm --workspace/--workspaces in the npm shim (#​1916, #​1917, @​jdx): the npm shim now translates these flags for run, test, start, stop, restart, and exec. Before, the flags reached the root script as arguments. A root script that called npm run build --workspaces then ran itself again and again. Like npm, --workspaces now skips the root unless you also pass --include-workspace-root. Arguments after -- are passed through unchanged.
    npm run build --workspaces     ->  aube -r --no-include-workspace-root run build
    npm run build -w @scope/child  ->  aube --filter @scope/child run build
    npm run build -w packages/child ->  aube --filter ./packages/child run build
    
  • npm --audit/--fund flags (#​1908, @​jdx): for install, add, ci, and remove and their aliases, the npm shim now drops --audit, --no-audit, --fund, --no-fund, and their =<bool> forms. Before, tools that pass these flags, such as the shadcn CLI, failed with unknown option '--no-audit'.
  • aube audit in workspaces (#​1927, @​jdx): aube audit now checks the direct dependencies of every workspace package, as pnpm audit does. It also finds the workspace lockfile when run from a member directory. Before, a workspace whose root had no dependencies printed No dependencies to audit. and exited 0. Running it from a member failed with "no lockfile found". --json dependency counts now add up all workspace packages. aube audit --fix and aube sbom still check only the root.
  • pnpm lockfile compatibility (#​1904, #​1914, @​jdx): aube no longer writes entries that pnpm rejects:
    • It no longer adds gitHosted: true to type: git resolutions, such as GitHub dependencies pinned to a short commit.
    • It no longer adds gitHosted: true to GitHub Packages (npm.pkg.github.com) tarballs. It removes this field from existing GitHub Packages entries when it rewrites the lockfile.
    • It now omits peerDependenciesMeta entries that are not optional, such as vitest's vite: {}. These entries made pnpm install --frozen-lockfile fail with ERR_PNPM_BROKEN_LOCKFILE.
  • Yarn Classic optionalDependencies (#​1922, @​jdx): aube now keeps the optionalDependencies: entries in yarn.lock when it reads and writes the file. Before, aube skipped platform binaries such as @rolldown/binding-darwin-arm64, so tools like vite 8 failed at runtime. Yarn 1 doesn't record os/cpu in the lockfile, so aube may install optional packages for every platform for now.
  • Removing a package that provides an optional peer (#​1920, @​Aschenbath): when you remove a direct dependency that provides an optional peer, aube now removes it from the lockfile and from hidden hoisting if nothing else needs it. Before, the package stayed resolvable through NODE_PATH. When aube reuses a lockfile, it also no longer treats a remembered optional-peer binding as a declared dependency.
  • latest with a minimum release age (#​1928, @​jdx): when minimumReleaseAge is set and the latest dist-tag points to a prerelease older than the cutoff, aube now resolves to that prerelease. Before, it chose the highest stable version, and minimumReleaseAgeExclude had no effect. This affects install, add, and outdated.
  • Vercel CLI with trustPolicy=no-downgrade (#​1910, @​jdx): vercel@>=58.5.1 is now in the default trustPolicyExclude. These releases are built from a private repository, so they have no provenance and failed the downgrade check. aube still checks earlier releases.
  • HTTP/2 too_many_data_frames errors (#​1912, @​jdx): some registries, such as npm.flatt.tech, send large package metadata in many small HTTP/2 frames, which caused the HTTP/2 client to close the connection. When this error occurs, aube now retries over HTTP/1.1 and keeps using HTTP/1.1 for metadata requests for the rest of the process.
  • Zero-padded tarball versions (#​1926, @​jdx): aube now accepts packages whose tarball package.json uses a zero-padded version, such as 2024.04.13, that matches the registry version 2024.4.13. Before, these installs failed with a content mismatch error.
Documentation
  • The installation guide now says that shell completions don't require a separate usage install. It also lists the supported shells: Bash, Zsh, Fish, PowerShell, Elvish, and Nushell (#​1915, @​jdx).

Full Changelog: aubepkg/aube@v2.9.0...v2.10.0

💚 Sponsor aube

aube is built and maintained by @​jdx, an open source developer at entire.io, the title sponsor of his open source work.

If aube saves your team install time or CI minutes, please consider becoming an individual or company sponsor. Your support funds ongoing development and helps keep aube fast, free, and independent.

docker/login-action

v4.8.0

Compare Source

Full Changelog: docker/login-action@v4.7.0...v4.8.0

oxc-project/oxc (oxfmt)

v0.78.0: oxfmt v0.78.0

Compare Source

🐛 Bug Fixes
  • ab78621 formatter_css/less: Print ~ escape consistently (#​29089) (Yuji Sugiura)
  • 91721d2 formatter_css/less: Print Less % selectors (#​29088) (Yuji Sugiura)
  • 449e91f oxfmt: Format ts-in-astro as TSX (#​29080) (leaysgur)
  • 8d8845e formatter/jsdoc: Strip * prefixes from multiline default value (#​28910) (leaysgur)
  • cdb1501 formatter/jsdoc: Handle every line terminator in multi-line JSDoc types (#​28909) (leaysgur)
  • fd51bff jsdoc: Handle every line terminators as newline (#​28902) (leaysgur)
  • 2f80765 formatter_toml: Format whitespace-only document to empty (#​28899) (leaysgur)
  • 5e1e02c formatter_graphql: Allow empty document (#​28898) (leaysgur)
  • 06e72c0 formatter/tailwind: Normalize CR in Tailwind template with preserveWhitespace (#​28896) (leaysgur)
⚡ Performance
  • 934ca98 formatter: Use Arc for tailwind options across embedding (#​29079) (leaysgur)
  • dde9c5c formatter_htmlish: Skip </script escape if possible (#​29078) (leaysgur)
  • dc4d918 formatter: Format embedded expr of one name without parsing (#​29077) (leaysgur)
📚 Documentation
  • eb2eb2d formatter_htmlish: Cover block elements in DIVERGENCES.md#interpolation-element-as-text (#​28901) (leaysgur)
  • 0f4ec80 formatter/tailwind: Note Tailwind sorting tears escaped whitespace apart (#​28900) (leaysgur)

v0.77.0: oxfmt v0.77.0

Compare Source

💥 BREAKING CHANGES
  • 00d1a67 oxfmt: [BREAKING] Format parser:vue files by oxc_formatter_htmlish (#​28848) (leaysgur)
🚀 Features
  • 9a81e58 formatter_htmlish: Format Vue templates (#​28847) (leaysgur)
  • 119d82f formatter: Format a template's expressions and handlers as JsRoot (#​28846) (leaysgur)
🐛 Bug Fixes
  • fa66fde oxlint,oxfmt: Resolve Vite+ config root from the config file directory (#​28801) (MK (fengmk2))
  • 4d4104d formatter_htmlish: Preserve doctype case in embedded (#​28797) (leaysgur)
  • ebfd40a formatter: Sort Tailwind calls printed as member chains (#​28800) (Yuji Sugiura)
  • e62e5cb formatter: Decide quotes of a Tailwind key printed as one class (#​28798) (Yuji Sugiura)
  • 959edc1 formatter: Don't sort templates in nested non-Tailwind calls (#​28794) (Yuji Sugiura)
  • 6ffbe32 formatter/jsdoc: Prevent @tag from being moved to the line start (#​28835) (leaysgur)
  • 94fbef5 formatter_css: Treat comments in style attr as fill items (#​28832) (leaysgur)
  • e4a80f7 formatter_css: Normalize selectors inside :deep, :slotted and :local (#​28805) (leaysgur)
  • 255c5cc formatter: Measure Tailwind class argument after collapsing whitespace (#​28745) (Lee Jaeha / Adam)
  • e6c4fef formatter: Decide template and object key layouts from collapsed Tailwind classes (#​28791) (Yuji Sugiura)
  • 0b44164 formatter: Decide JSX opening element layout from collapsed Tailwind class string (#​28744) (Lee Jaeha / Adam)
  • b8ba9ae formatter: Print decorators on rest parameters (#​28731) (YT)
⚡ Performance
pullfrog/pullfrog

v0.1.129

Compare Source

This release limits what a run started by someone without write access can do. On a public repo, that includes any outsider.

[!IMPORTANT]
A run whose triggering user has read or no access to the repo can now write only to the issue or pull request it was started on. Runs started by collaborators (triage and above) and runs with no triggering user (autofix, dispatch) are unchanged.

Security

Area What changed
Writes to other threads A reader-triggered run can comment, reply to review comments, review, resolve threads and update the PR body only on its own issue or PR. It can edit only Pullfrog's comments there. It can add labels only to its own issue, never a PR, and it can't remove labels, open issues or open pull requests
@pullfrog in an outsider's issue, PR or review Only an author allowed by the repo's Mentions setting turns the body into the run's prompt, with its --model, --effort and --notimeout flags. Anyone else's text is context, and the repo's own instructions run
Fork pushes A push to the fork head of a collaborator's PR is re-reviewed at the lower of the PR author's and the pusher's roles

All three are in 791864e.


Full Changelog: pullfrog/pullfrog@v0.1.128...v0.1.129

v0.1.128

Compare Source

This release fixes a batch of run failures found by the run and log audits, from harness installs that left a placeholder binary to OpenCode startups that hung silently.

[!NOTE]
A run that fails because it landed no GitHub write now turns the pullfrog check red. It used to leave the check green.

Fixes

Area What changed
Harness install An install that leaves the package's placeholder binary in place is retried once from an empty tree, then reported with npm's output, instead of failing with claude native binary not installed. The install passes --include=optional --ignore-scripts=false, so a host .npmrc can't skip the platform binary
OpenCode startup Startup gets 180s, then one retry on a fresh server, instead of a silent 15-minute hang. opencode serve gets 120s to bind instead of 30s, and opencode models retries OpenCode V2's plugin refusal
Proxied runners Codex and ask_model get the runner's HTTPS_PROXY, NO_PROXY and related variables, so Codex reaches its model on a proxied runner. The agent's shell still doesn't see them
pullfrog check The check and auto-merge are finalized after the run's final outcome
Reviews When GitHub refuses a verdict for lack of explicit repository access, the review posts as a comment. A turn that dies after its review posted reports success. An inline comment with no body and no suggestion is refused before anything reaches GitHub
Root runners Claude Code runs on runners that use uid 0, via IS_SANDBOX=1
Auto-select A run with only CODEX_AUTH_JSON gets an OpenAI model
Startup run-context retries a 5xx or a dropped attempt twice
pnpm The corepack fallback pins pnpm to the action's own version instead of latest, which older corepacks can't launch
Partial clones checkout_pr and the incremental diff fetch missing objects with credentials first, so git's credential-less lazy fetch never fires
Keys A key holding a smart quote or other non-ASCII character is refused with a message naming it, instead of failing as a ByteString error
Progress comment A progress comment deleted during a run falls back to a new comment. A cancel recorded before the run was claimed reaches GitHub
Codex plans Codex plans prolite and pro are labelled "Pro", and promax "Pro Max"
Paused organizations The paused-organization notice now says why Pullfrog paused and what the Organization plan includes

The audit fixes are in 2416011, the Codex plan labels in 7102c8a, and the paused-organization notice in ec2697f.


Full Changelog: pullfrog/pullfrog@v0.1.127...v0.1.128

v0.1.127

Compare Source

This release adds Google's Gemini 4 Carbon.

Features

Area What changed Link
Gemini Carbon gemini-carbon runs Gemini 4 Carbon on the google, opencode, openrouter and vercel routes. It sits beside gemini-pro, which stays on Gemini 4 Argon. Review subagents drop to gemini-pro 553b20b

Internal

  • The progress comment's lifecycle is an explicit state. (08b8438)
  • The checkout-moved self-test tells its agent what the swap script is, so models stop refusing it as a prompt injection. (513e4c7)

Full Changelog: pullfrog/pullfrog@v0.1.126...v0.1.127

v0.1.126

Compare Source

This release moves Claude Fable to Fable 5.2 and adds Meta's Muse Wonder and Muse Ember.

Features

Area What changed Link
Claude Fable anthropic/claude-fable and opencode/claude-fable now run Claude Fable 5.2. The Claude Code harness moves to 2.1.321, the first version that accepts the model 43ec500
Muse Wonder and Muse Ember Meta's top tier (muse-wonder, Muse Wonder 1.0) and its cost-efficient tier (muse-ember, Muse Ember 1.0) join Muse Spark on the meta, opencode, openrouter and vercel routes 20615eb

Full Changelog: pullfrog/pullfrog@v0.1.125...v0.1.126

v0.1.125

Compare Source

This release lets a run ask another model for a second opinion, using any model your account holds a credential for.

Features

Area What changed Link
ask_model tool A run can hand a question to another model, with an optional effort, and use the answer. For example, a prompt can ask a second model to try to refute each review finding before it posts. The other model has the run's tools except those that push or post. Each run gets 12 calls of up to 9 minutes each, and the tool appears only when the run can reach another model. See Asking other models 0cc162f

Full Changelog: pullfrog/pullfrog@v0.1.124...v0.1.125

v0.1.124

Compare Source

This release recovers runs whose first model call is refused on a Pullfrog-issued key, and stops reviews from running out of time on re-dispatched specialists.

Fixes

Area What changed Link
OSS and Router runs A first model call refused with 401 User not found on a key Pullfrog issued for the run now waits 15s, then 30s, and tries again in the same session. If the refusal persists, the failure says it is Pullfrog's credential instead of telling you to rotate a key you never held 229b3b5
Reviews near their time limit A review specialist stopped for its time budget is no longer dispatched again, because the run is now told why it stopped. Runs entering their last 5 minutes are told to submit what they have, and new specialists are refused from then on 229b3b5

Full Changelog: pullfrog/pullfrog@v0.1.123...v0.1.124

v0.1.123

Compare Source

This release fixes web search on OpenCode V2 and shows a provider's own error text on failed runs.

Fixes

Area What changed Link
Web search on OpenCode V2 Runs no longer wait a minute on a consent form and fail with Web search cancelled. opencode/* and opencode-go/* models search through Exa, as they did on V1. Other models don't get the tool unless the repo sets a provider in its own opencode.json a9022b4
Tools that ask for input A tool that asks for a person's input, such as question or an MCP server's elicitation, is refused at once instead of waiting on an answer that can't come a9022b4
Provider errors The provider's own error text now reaches the log and the PR comment, in place of a bare Invalid openai-compatible/openai-compatible-chat stream event. Zen's free usage limit gets the out-of-credit message again a9022b4

Full Changelog: pullfrog/pullfrog@v0.1.122...v0.1.123

v0.1.122

Compare Source

This release lets a run send custom request headers to a gateway or OpenAI-compatible endpoint, and retires the withdrawn Pulse Alpha model.

Features

Area What changed Link
Custom request headers Each endpoint takes its own headers variable, one Name: Value per line: OPENAI_COMPATIBLE_HEADERS for an unnamed OpenAI-compatible endpoint, OPENAI_COMPATIBLE_<NAME>_HEADERS for a named one, OPENAI_CUSTOM_HEADERS with an OPENAI_BASE_URL gateway (OpenCode and Codex), and ANTHROPIC_CUSTOM_HEADERS with an ANTHROPIC_BASE_URL gateway, which OpenCode now reads as Claude Code already did. Headers go only to their own endpoint, values are masked in logs, and a malformed line fails the run before the agent starts with Line N of <VAR> is not a header. The console's provider cards take them under Send custom headers 70654c6

Fixes

Area What changed Link
Pulse Alpha OpenCode Zen withdrew the free opencode/pulse-alpha model. A repo that selected it now runs on Big Pickle b74841a
Runtime prompt The prompt no longer carries an unused shellUnsandboxable line on every run 79c6e5d

Full Changelog: pullfrog/pullfrog@v0.1.121...v0.1.122

v0.1.121

Compare Source

This release fixes the to-do spinner on GitHub's new pull request Overview page, a false failure on reviews posted outside the review tool, and tells the agent why it has no shell on some self-hosted runners.

Fixes

Area What changed Link
To-do spinner The spinner no longer renders at full comment width on GitHub's new pull request Overview page. It is now a sharp 12px SVG instead of a 150x150 GIF dfa0da5
Review and IncrementalReview A Pullfrog review submitted during the run now counts as delivered even when create_pull_request_review did not post it. Those runs used to post a second review or end red with Review mode finished without calling create_pull_request_review. Fixes #​127 bdefd51
Self-hosted runners without a sandbox The agent is now told why it has no shell. When that leaves its result unverified, such as a fix whose tests it could not run, its final comment or review says so and links the runner setup docs 8c63c09

Full Changelog: pullfrog/pullfrog@v0.1.120...v0.1.121

[v0.1.120](https://redirect.github.com/pullfrog/pul

[!IMPORTANT]
✂ PR body was truncated to here.


Configuration

📅 Schedule: (UTC)

  • Branch creation
    • "before 12pm on Sunday"
  • Automerge
    • At any time (no schedule defined)

🚦 Automerge: Disabled by config. Please merge this manually once you are satisfied.

♻ Rebasing: Whenever PR becomes conflicted, or you tick the rebase/retry checkbox.

👻 Immortal: This PR will be recreated if closed unmerged. Get config help if that's undesired.


  • If you want to rebase/retry this PR, check this box

This PR was generated by Mend Renovate. View the repository job log.

@renovate
renovate Bot requested a review from RedStar071 as a code owner November 15, 2026 02:27
@coldtea-pr-lens

Copy link
Copy Markdown

Note

renovate[bot] opened this pull request, so PR Lens left it undrawn. Comment @pr-lens draw to draw it

github.comment.notice: false in .github/pr-lens.yml turns this note off

@socket-security

socket-security Bot commented Nov 15, 2026 •

Copy link
Copy Markdown

Review the following changes in direct dependencies. Learn more about Socket for GitHub.

Diff Package Supply Chain
Security
Vulnerability Quality Maintenance License
Updatedoxfmt@​0.76.0 ⏵ 0.78.0941008896 +2100

View full report

@renovate
renovate Bot force-pushed the renovate/all-minor-patch branch from 7c049e2 to 1071256 Compare November 16, 2026 02:15
@RedStar071
RedStar071 merged commit dd07e1a into main Nov 16, 2026
5 checks passed
@RedStar071
RedStar071 deleted the renovate/all-minor-patch branch November 16, 2026 15:59
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant