The multi-line git commit message inside the run block dropped below the
literal block's indentation, ending the block mid-command and failing the
whole workflow file at parse time (0s failure, no jobs). Use two -m flags on
one shell line instead. YAML now validated locally before push.
Token included so this push performs the regeneration.
Co-Authored-By: Claude Fable 5 <[email protected]>
Claude-Session: https://claude.ai/code/session_01LHH3wPBsoruwCaJo8Rpf6a
workflow_dispatch only works for workflows on the default branch, which this
file will not reach until the next promotion. Gate the job on an explicit
[update-visual-baselines] commit-message token instead so it is dispatchable
from unstable now; the workflow's own commit omits the token, so it cannot
retrigger itself.
This commit carries the token, so pushing it performs the first regeneration.
Co-Authored-By: Claude Fable 5 <[email protected]>
Claude-Session: https://claude.ai/code/session_01LHH3wPBsoruwCaJo8Rpf6a
The visual-regression job pixel-compares tests/visual/baselines.json on
ubuntu-latest, but the committed baselines were captured on developer machines.
Chromium builds and font fallbacks differ between environments — the token
stack's "Inter"/"Fira Code" are not shipped or installed anywhere, so every
machine renders its own system fallback — and the kitchen-sink default-theme
diff drifted past MAX_DIFF_RATIO on CI while passing 13/13 locally. ci.yml
documents exactly this mismatch as the reason the job is quarantined from
blocking deploys.
The CI diff artifact confirms it: baseline and actual render identically except
for ghosted/offset text in the high-contrast lower half — font-metric drift,
not a design change. (The dark upper half diffs too little to trip the 0.2
pixelmatch threshold, which is why only the default theme fails.)
Fix the class of problem instead of the instance: capture baselines in the SAME
runner image, browser install and font stack that the comparison job uses.
workflow_dispatch only — regenerating baselines is an explicit decision made
when a visual change is intentional, never a side effect of a push. The
workflow re-runs the comparison against its own output before committing, so a
bad capture cannot land.
Co-Authored-By: Claude Fable 5 <[email protected]>
Claude-Session: https://claude.ai/code/session_01LHH3wPBsoruwCaJo8Rpf6a
The v5 line of these two actions still targets Node.js 20 (unlike
checkout/setup-node@v5, which are already on Node 24), so the deprecation
warning persisted. upload-artifact reaches Node 24 at v6+ and
download-artifact at v7+; move to the current latest (v7 / v8), which
share the same v4-generation artifact backend and pair correctly.
Co-Authored-By: Claude Opus 4.8 <[email protected]>
Claude-Session: https://claude.ai/code/session_01DwbE3uTiFaKRNjx5vP2muV
The v4 actions target Node.js 20, which GitHub now force-runs on Node 24
and flags as deprecated. Bump all four across every workflow to v5 (their
native Node-24 line) to clear the warning. Mechanical version-only change;
upload/download-artifact stay on matching majors as required.
Co-Authored-By: Claude Opus 4.8 <[email protected]>
Claude-Session: https://claude.ai/code/session_01DwbE3uTiFaKRNjx5vP2muV
Add a `release` job (main only, contents: write) that builds the archive
and publishes it as an immutable GitHub Release tagged
site-<date>-<shortsha>, marked --latest, with bitvid-site.zip attached.
This gives a stable download URL for the current production archive:
https://github.com/PR0M3TH3AN/bitvid/releases/latest/download/bitvid-site.zip
Re-runs of the same commit refresh the asset instead of failing. The
Release holds only the public compiled dist/ (same as Vercel serves) and
no Nostr signing key — publishing to nsite stays a manual, client-signed
step. Per-branch artifact upload is unchanged. Docs updated.
Co-Authored-By: Claude Opus 4.8 <[email protected]>
Claude-Session: https://claude.ai/code/session_01DwbE3uTiFaKRNjx5vP2muV
Add .github/workflows/nsite-archive.yml (push to any branch +
workflow_dispatch): npm ci + npm run build, then upload dist/ as an
artifact (bitvid-site-<branch>-<shortsha>, 30d retention). GitHub serves
it as a .zip with index.html at the root, ready for a manual nsite
publish. The workflow holds no signing key and publishes nothing.
Also add an `archive` npm script for the same output locally
(bitvid-site.zip, now gitignored), and land docs/nsite-mirror-plan.md
revised to the manual-publish + auto-archive model: sign client-side via
NIP-07 or local nsyte, never an nsec/nbunksec in CI. The prior CI-signing
workflow is retained as a labeled Phase-3 reference option.
Co-Authored-By: Claude Opus 4.8 <[email protected]>
Claude-Session: https://claude.ai/code/session_01DwbE3uTiFaKRNjx5vP2muV
The visual-regression step lived inside the build job, and deploy-cache-state
(the Cloudflare CDN cache-purge that makes new content visible) needs: build.
So an environment-sensitive pixel diff — CI's Chromium/font rendering differs
from the committed baseline (locally the baseline is byte-identical, so it's not
a content regression) — failed the build job and silently blocked every deploy.
The unstable.bitvid.network CDN was stuck on the 2026-06-17 build as a result.
Move the visual steps into a standalone visual-regression job so they still run
and report a failure, but no longer block the build job or the deploy. Real
lint/build/artifact failures still gate the deploy via build.
Co-Authored-By: Claude Opus 4.8 <[email protected]>
- Update .github/workflows/ci.yml to install firefox for e2e-headless job
- Update js/adminListStore.js to await nostrClient.initPromise if pool is not ready, preventing race conditions during app startup
Co-authored-by: PR0M3TH3AN <[email protected]>
Explicitly injects the GITHUB_TOKEN into the upstream remote URL in the agent-pr-review workflow. This prevents authentication failures when fetching the upstream repository in CI environments where plain HTTPS URLs trigger interactive credential prompts.
Co-authored-by: PR0M3TH3AN <[email protected]>
- Add docs/capacitor-native-app-plan.md to hex-lint ignore list (code examples
naturally reference hex colors in Capacitor config)
- Replace direct console.error/console.log calls with logger utility in
historyView.js, profileCache.js, EventDetailsModal.js, ZapInterface.js,
and notificationController.js
- Update CI workflow from Node 20 to Node 22 to match package.json engines
https://claude.ai/code/session_01LzErpH9FNw4hAPsjoYb1v9
Replaces `git rev-parse` with `git ls-remote` to verify the existence of the upstream base branch before attempting to fetch and diff. This prevents workflow failures when the PR targets a branch that does not exist on the upstream remote (e.g. `unstable` or other feature branches), gracefully skipping the diff step with a warning instead.
Co-authored-by: PR0M3TH3AN <[email protected]>
Updates the `agent-pr-review.yml` workflow to prevent failures when calculating diffs against the upstream repository.
Changes:
- Reverts `git fetch` to a full fetch to ensure all remote-tracking branches (like `refs/remotes/upstream/unstable`) are correctly updated in the local environment.
- Adds a mandatory `Verify upstream branch and diff` step that checks for the existence of the target base branch (`upstream/${{ github.base_ref }}`) using `git rev-parse`.
- Performs a dry-run `git diff` to confirm connectivity before running the main review script.
- Ensures the job fails explicitly with a clear error message if the upstream branch is missing, rather than crashing later with an obscure exit code 1.
Co-authored-by: PR0M3TH3AN <[email protected]>
- Add `scripts/automated-pr-review.mjs` to run format, lint, tests, and security checks.
- Add `.github/workflows/agent-pr-review.yml` to trigger the review on PRs.
- Script flags sensitive file changes (crypto/storage) and posts comments.
- Script attempts to commit formatting micro-fixes if permissions allow.
