Add path filtering to TS workflows to skip them for Elixir-only changes

[alco]

Summary

• Skip TS workflows on PRs that only touch Elixir packages, integration tests, or READMEs
• Add changeset-release/main to push triggers
• Add push trigger to ts_check_formatting.yml for main and changeset-release/main
• Pushes to main and changeset-release/main always run TS CI (no path filtering)

Approach

Uses paths-ignore to exclude the few non-TS paths rather than an allowlist of every TS-related path. Since the repo is majority TS, the ignore list is shorter and doesn't need updating when new TS packages or config files are added.

Ignored paths:

• packages/sync-service/**, packages/electric-telemetry/**, packages/elixir-client/**
• integration-tests/**
• **/README.md
• website/** (in ts_tests.yml only — formatting covers website)

Branch protection note

GitHub treats workflows skipped due to path filtering the same as workflows where all jobs are skipped — required status checks from these workflows will not block merge for Elixir-only PRs.

Test plan

• Verify this PR triggers TS CI (workflow files are not in the ignore list)
• Verify an Elixir-only PR does not trigger TS CI
• Verify pushes to main always trigger TS CI

🤖 Generated with Claude Code

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 4ea2695a38

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 88.67%. Comparing base (0b45e63) to head (43b165e).
✅ All tests successful. No failed tests found.

Additional details and impacted files

@@           Coverage Diff           @@
##             main    #4069   +/-   ##
=======================================
  Coverage   88.67%   88.67%           
=======================================
  Files          25       25           
  Lines        2438     2438           
  Branches      613      613           
=======================================
  Hits         2162     2162           
  Misses        274      274           
  Partials        2        2           

Flag	Coverage Δ
packages/experimental	87.73% <ø> (ø)
packages/react-hooks	86.48% <ø> (ø)
packages/start	82.83% <ø> (ø)
packages/typescript-client	93.81% <ø> (ø)
packages/y-electric	56.05% <ø> (ø)
typescript	88.67% <ø> (ø)
unit-tests	88.67% <ø> (ø)

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[claude]

Claude Code Review

Summary

This PR adds path filtering to TS CI workflows so they are skipped for Elixir-only PRs. Since the previous review, the implementation was simplified from an allowlist (paths) approach to a denylist (paths-ignore) approach. The change is well-scoped and correct.

What's Working Well

• Simplifying to paths-ignore (denylist) is a good call: new TS packages added in future will automatically trigger CI without needing to be added to an explicit allowlist.
• Push triggers now have no path filtering at all — all pushes to main and changeset-release/main run TS CI, which is the right trade-off (lower risk of a regression slipping through on merge).
• Adding push trigger to ts_check_formatting.yml closes the gap where formatting regressions could merge undetected.
• Adding changeset-release/main to the ts_tests.yml push trigger closes a similar gap for release branches.
• The packages/sync-service/** exclusion concern raised by the bot was correctly dismissed by the author: sync-service changes will still be caught by push-to-main CI.
• The .agent-tasks entry in .prettierignore is harmless cleanup.

Issues Found

Critical (Must Fix)

None.

Important (Should Fix)

None.

Suggestions (Nice to Have)

• Flaky test still present: The Codecov report still shows test/client.test.ts > should fall back to long polling after 3 consecutive short SSE connections failing (AssertionError: expected 5 to be less than or equal to 4). This is unrelated to these workflow changes — worth confirming it's a known flaky test tracked elsewhere before merging.

Issue Conformance

No linked issue. Acceptable for a CI housekeeping change; the PR description explains the motivation clearly.

Previous Review Status

• Resolved: packages/start/ coverage concern — no longer applicable with the denylist approach.
• Resolved: Allowlist completeness concerns — not applicable with the denylist approach.
• Remaining: Flaky test noted in previous review is still present in CI output (unrelated to this PR).

---

Review iteration: 2 | 2026-03-27