v0.23.0 still panics with SendError (check.rs:209/165) in GitHub Actions; request stable release with #2074

[AccessiT3ch]

Summary

lychee-v0.23.0 still crashes in CI with tokio SendError panics instead of reporting a normal link/network error.

I can see that #2074 was merged and appears to address this class of panic, but v0.23.0 is still the latest stable, so GitHub Action users are still hitting hard failures.

Environment

• lycheeverse/lychee-action@v2
• action default tool version: lychee-v0.23.0
• GitHub Actions runner (ubuntu-latest)

Observed failure

From CI logs:

thread 'tokio-runtime-worker' panicked at lychee-bin/src/commands/check.rs:209:18:
cannot send response to queue: SendError { .. }

thread 'main' panicked at lychee-bin/src/commands/check.rs:165:38:
Cannot send request: SendError { .. }

This exits non-zero and blocks required checks.

Why this seems related to #2074

The merged PR #2074 describes the same SendError/tokio panic class and changes behavior to report actionable errors rather than panic.

Related historical issues

• Lychee don't panic, invalid URL's are fine. #539 (Lychee don't panic, invalid URL's are fine.)
• Error: thread 'tokio-runtime-worker' panicked at 'not currently running on the Tokio runtime.' #131 (older tokio runtime panic class)

Request

1. Can maintainers confirm whether Report input source errors rather than tokio panics #2074 fully fixes this v0.23.0 panic path?
2. If yes, can we get a stable release that includes Report input source errors rather than tokio panics #2074 soon (e.g., v0.23.1 / next patch)?
3. If not, what additional repro detail would help isolate this specific crash path?

I’m happy to provide full failing run logs if useful.

[katrinafyi]

Help me out - can you try quoting 'docs/**/*.md' in your lychee args and see if the issue with whitespace files persists? It's probably a shell quoting issue.

Also, you can also try lycheeVersion: nightly in the action to unblock your velocity. It would also help me to know if it fixed the issue for you.

[SevWren]

Help me out - can you try quoting 'docs/**/*.md' in your lychee args and see if the issue with whitespace files persists? It's probably a shell quoting issue.

Also, you can also try lycheeVersion: nightly in the action to unblock your velocity. It would also help me to know if it fixed the issue for you.

Quick field report from another affected repo.

I hit the same Lychee crash pattern in GitHub Actions in my repository’s Docs Link Check workflow (using lycheeverse/lychee-action@v2).`

The failure matched the SendError panic signature reported in this issue.

Applying the suggested workflow changes ('docs/**/*.md' + lycheeVersion: nightly) stopped the panic path and surfaced normal actionable link results. docs-link-nightly-workaround.yaml

Environment

• Workflow: Docs Link Check (docs-link)
• Provider: GitHub Actions (ubuntu-latest)
• Action: lycheeverse/lychee-action@v2
• Failing config at time of panic:
  • args: --config .lychee.toml README.md docs/**/*.md

Observed failure

• Workflow run failed with panic and exit code 101.

From CI logs:

thread 'tokio-runtime-worker' panicked at lychee-bin/src/commands/check.rs:209:18:
cannot send response to queue: SendError { .. }
note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace
thread 'main' panicked at lychee-bin/src/commands/check.rs:165:38:
Cannot send request: SendError { .. }
Error: Process completed with exit code 101.

[katrinafyi]

Y'all are so wordy lol, but it's good to see that the fix worked. I'd be interested to know what the real error was after the panic was removed.

Quoting the glob was my guess at fixing a root cause based on @ AccessiT3ch's linked issue. It may or may not be needed if your error was something else.

Anyway, I don't think anything more needs to be done for this ticket. The release schedule has been discussed before in #2025.