cargo update -p rustls-webpki -p tar moves rustls-webpki 0.103.9 -> 0.103.13
and tar 0.4.44 -> 0.4.46, both within the existing Cargo.toml bounds. They sit
under tauri-plugin-updater, which downloads and unpacks application updates, so
all six of their advisories were reachable in the shipped binary.
The remaining three can neither be fixed nor reached. quick-xml (2x 7.5 high)
is pulled by plist, which tauri only needs for Apple bundling: its per-target
trees are empty for both shipped targets and it appears solely under
x86_64-apple-darwin. Its fix is >= 0.41.0 while plist requires ^0.38, a
semver-incompatible boundary [patch.crates-io] cannot cross. rsa has no
published fix at all and is never compiled — its only parent is sqlx-mysql, an
artifact of sqlx's multi-backend graph on a SQLite project.
Leaving those three to red the daily gate forever would reproduce the signal
loss that #232 removed the `|| true` to fix, so they move into a versioned
.cargo/audit.toml. Entries are keyed by advisory ID, never by crate, so a new
advisory against the same crate still reds the gate; each carries its
reachability proof and its removal condition.
A blocking step in check-rust.yml re-proves that justification on every PR
touching src-tauri/ or .cargo/, and fails if a suppressed crate enters a
shipped target's graph — the scenario that would rot the list is itself a
src-tauri change. It separates cargo tree's exit status from its output (an
absent crate and a failed invocation both print nothing) and asserts a canary
crate is still found, so its silence proves something.
cargo audit: 9 vulnerabilities -> 0, warnings unchanged at 23
(cargo-audit 0.22.2, advisory-db 0bfde9d6 of 2026-07-27).
cargo check + cargo test green (106 tests); npm build + 871 vitest green.
Resolves#310
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Follow-up on the review of #232:
- .claude/ is tracked (rules + skills) and cannot affect the frontend build,
so a change confined to it no longer queues a 1m43s job for nothing.
- The release skill's pre-flight step named check.yml, which this PR deletes.
The reasoning still holds — the check workflows only run on PRs, never on
main, so a merged tip has never been seen by CI — only the filename moved.
Its dated changelog entry is left alone.
Resolves#232
The rust job cost 21m44s on every PR while only ~1 PR in 40 touches
src-tauri/, and the runner has capacity 1 — the frontend job queues behind
it, so every PR paid ~24.5 min of feedback.
Measured on run 326 (2026-07-21), 12m15s of that was pure waste:
- 6m54s tarring target/ and the cargo registry for saves that time out
against the runner's unreachable cache server (#234). The restore times
out into a miss too, so nothing was ever cached at either end.
- 4m41s recompiling cargo-audit from source on every run.
- ~40s on the two doomed restores.
Split check.yml into check-rust.yml (paths: src-tauri/**) and
check-frontend.yml (paths-ignore denylist), drop every actions/cache step
until #234 is fixed, and install cargo-audit as a prebuilt binary via
taiki-e/install-action. The audit step keeps continue-on-error — advisories
are informational and can land on unrelated crates — but loses the `|| true`
that also hid tooling failures; the install step is blocking.
The frontend filter is a denylist on purpose: that job costs ~2.5 min, so
running it needlessly is cheap while silently not running it is not. The
expensive job keeps a strict allowlist.
Neither workflow filters on `branches:` anymore. `branches: [main]` never
matched a PR stacked on another feature branch, which is what /autopilot
produces: PRs #305-#308 of the feature-gating milestone ran no CI at all.
Adds audit.yml for daily RustSec coverage, since check-rust.yml now only
runs on Rust PRs. It skips the Rust toolchain entirely — cargo-audit only
reads Cargo.lock — so it costs ~1-2 min rather than the ~22 a scheduled
check-rust would burn daily on a capacity-1 runner.
The GitHub mirror is left untouched (#233: it receives no PRs).
Expected: Rust PR ~9-10 min, frontend-only PR ~2.5 min instead of ~24.5.
Resolves#232
The `push` + `pull_request` combo doubled CI runs on every PR (visible
on #170 with 4 pending checks instead of 2). Drop `push`, keep
`pull_request: branches: [main]`.
Trade-off: branches pushed without an open PR no longer get CI
feedback. Open a draft PR if you want CI to run before requesting
review — `/fix-issue` always opens a PR right after pushing, so the
gap is essentially zero in practice.
Also adds a concurrency group `ci-${{ github.ref }}` with
cancel-in-progress so force-pushes cancel the previous run instead
of stacking.
Same change applied to .github/workflows/check.yml (GitHub mirror)
to keep the two configs in sync.
Fixes#171
The new token_store module (#78) depends on `sync-secret-service` via
`dbus-secret-service`, which in turn links to libdbus-1 at build time
through the `dbus` crate. Add `libdbus-1-dev` to:
- `check.yml` rust job (alongside the existing webkit/appindicator
system deps), so every PR run compiles the keyring backend.
- `release.yml` Linux deps step, so tagged builds link correctly.
Runtime requires `libdbus-1-3`, which is present on every desktop
Linux distro by default, so `.deb` / `.rpm` depends stay unchanged.
Also add a non-blocking `cargo audit` step to check.yml to surface
advisories across the transitive dep graph (zbus, dbus-secret-service,
etc.) without failing unrelated PRs.
Drop `appimage` from `bundle.targets` in tauri.conf.json: the release
workflow explicitly builds `--bundles deb,rpm` so AppImage was never
shipped, and its presence in the config risks a silent fallback to
plaintext token storage for anyone running `tauri build` locally
without libsecret/libdbus bundled into the AppImage. No behaviour
change for CI; follow-up to re-enable AppImage properly would need a
linuxdeploy workflow that bundles the backend.
Refs #66
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
actions/checkout@v4 and actions/cache@v4 are JavaScript actions and
require `node` in the container PATH. The rust job in check.yml only
installed system libs and the Rust toolchain, so the post-checkout
cleanup failed with `exec: "node": executable file not found in $PATH`
on every Forgejo run.
The frontend job already installed Node, which is why it succeeded.
The GitHub mirror is unaffected because ubuntu-latest ships with Node
preinstalled.
Validated against the failed run https://git.lacompagniemaximus.com/maximus/Simpl-Resultat/actions/runs/122
Adds .forgejo/workflows/check.yml (and a GitHub mirror) that runs on
every branch push (except main) and on every PR targeting main.
Two parallel jobs:
- rust: cargo check + cargo test, with cargo registry/git/target caches
keyed on Cargo.lock. Installs the minimal Rust toolchain and the
webkit2gtk system deps that the tauri build script needs.
- frontend: npm ci + npm run build (tsc + vite) + npm test (vitest),
with the npm cache keyed on package-lock.json.
The Forgejo workflow uses the ubuntu:22.04 container pattern from
release.yml. The GitHub mirror uses native runners (ubuntu-latest)
since the GitHub mirror exists for portability and uses GitHub-native
actions.
Documents the new workflow in CLAUDE.md alongside release.yml so future
contributors know what CI runs before merge.
- Create CHANGELOG.fr.md with full French translation of all versions
- Add ChangelogPage (/changelog) accessible from Settings
- Replace hardcoded i18n release notes with file-based approach:
loads CHANGELOG.md or CHANGELOG.fr.md based on user language
- CI copies changelogs to public/ during release build
- Update CLAUDE.md with bilingual changelog deployment process (fixes#10)
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
The automatic GITHUB_TOKEN does not have package permissions in Forgejo.
Use a dedicated PACKAGE_TOKEN secret with the package scope instead.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
The DELETE endpoint for Forgejo generic packages is /api/v1/packages/
while the upload (PUT) endpoint is /api/packages/. Using the wrong
endpoint caused silent failures, leaving stale latest.json in the
registry. Also fail the CI step explicitly on upload failure.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Forgejo generic package API uses /api/packages/ not /api/v1/packages/
for both upload and deletion.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Prevents "Release has no Tag" error when re-running a build for
a tag that already has a release.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
The DELETE was using the upload URL instead of the API v1 package
endpoint, so old versions were never actually deleted, causing 409
on re-upload and leaving stale latest.json.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
The ${{ steps.changelog.outputs.notes }} expression was being interpolated
directly into shell commands, causing parse errors when the changelog
contained special characters, markdown, or backticks.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Fix updater: strip v-prefix from version in latest.json, and delete
old package before re-uploading to avoid 409 conflicts.
Add frontend log capture (console intercept) with a log viewer card
in the settings page (filterable, copyable, clearable).
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
linuxdeploy cannot run inside Docker containers even with
APPIMAGE_EXTRACT_AND_RUN. Use --bundles deb,rpm instead.
Update latest.json to use .deb for linux-x86_64 platform.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
linuxdeploy needs FUSE which is unavailable in containers.
This env var makes it extract instead of mounting.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Docker container steps don't share shell state. Set PATH at job level
to include /root/.cargo/bin and remove per-step source cargo/env.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
The Forgejo runner's default image uses glibc 2.39, which produces
binaries incompatible with Pop!_OS / Ubuntu 22.04 (glibc 2.35).
Build inside container: ubuntu:22.04 with Node.js and Rust installed.
Bump to v0.4.4.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Forgejo has no /releases/latest/download/ route (GitHub-specific)
- Upload latest.json to generic package registry for a stable endpoint
- Fix Linux signature collection: use .AppImage.sig (not .tar.gz.sig)
- Collect all platform signatures (.deb.sig, .rpm.sig, .AppImage.sig)
- Bump to v0.4.3
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Tauri CLI filters --bundles values by host OS. The nsis target is already
configured in tauri.conf.json and auto-selected for Windows targets.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Add Windows cross-compile (cargo-xwin) to Forgejo CI workflow
- Add libsqlite3-sys bundled for cross-compile compatibility
- Switch updater endpoint from GitHub to self-hosted Forgejo
- Collect Windows NSIS assets and include windows-x86_64 in latest.json
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Tauri requires TAURI_SIGNING_PRIVATE_KEY_PASSWORD to be defined,
even when the key has no password.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Replace dtolnay/rust-toolchain, actions/setup-node, and swatinem/rust-cache
with shell commands since they are not mirrored on code.forgejo.org.
Use https://github.com/ prefix for checkout action.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Add .forgejo/workflows/release.yml for Forgejo Actions
- Update signing pubkey for new key pair
- Sync Cargo.toml version to 0.4.0
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
