OpenClaw v2026.5.25-beta.1: Installer, operator field notes

The source record for v2026.5.25-beta.1 names a release-level runtime or packaging change. This is an operator's reading of that record, with the exact source facts kept next to a bounded experiment rather than a generic feature list.

Because this is a pre-release, use a disposable Gateway first and leave the stable version available for comparison. The release body cannot certify every local combination of provider, channel, platform, and account; use the bounded test below for the path you actually run.

A small field test

01 / Installer

Source signal. let the local-prefix CLI installer use Alpine's apk Node.js, npm, and Git packages on musl Linux instead of downloading glibc Node tarballs that fail node:sqlite.

The data boundary in Installer is clearer with two fixtures: one known-good input and one that should never be read. Verify both outcomes without using private material.

02 / Checks

Source signal. prefilter tracked conflict-marker scans so changed checks avoid reading every repository file on clean runs.

Record the source path, transformed output, and refusal reason when testing Checks; a successful preview alone does not prove that the boundary is enforced.

The details that affect a runbook

Read the narrow change in context

v2026.5.25-beta.1 is a pre-release checkpoint with a deliberately small surface. The source names Installer: let the local-prefix CLI installer use Alpine's apk Node.js, npm, and Git packages on musl Linux instead of downloading glibc Node tarballs that fail node:sqlite. A narrow patch still has a clear job: it should remove one failure or preserve one compatibility promise without changing unrelated configuration.

Keep this patch beside the operator's maintenance notes. Compare the installed tag, package version, and service owner before applying it. If this entry is a correction release, record the broken behavior that led you here, then repeat that exact scenario after the update. That evidence is more useful than a general claim that the package installed successfully.

The adjacent source detail is Checks: prefilter tracked conflict-marker scans so changed checks avoid reading every repository file on clean runs. Read both signals together, because a hotfix can alter the boundary around a plugin, channel, session, or installer even when the version number looks like a minor suffix.

A small worksheet for this version

• Write the old version, the new version, and the package or tag that actually changed.
• Capture one reproduction before the update and the same reproduction after it.
• Record the first useful log line, the final user-visible result, and any rollback command that was tested.
• Leave unrelated channels, skills, and provider defaults untouched until the correction is understood.

A bounded check for this version

1. Start with a disposable checkpoint: record the version, host, provider, and workspace tied to Installer.
2. Check the release's migration notes and make a restorable backup before changing persistent state.
3. Use one deterministic input to exercise the change and keep the command, output, and source note together.
4. Compare the result with the previous version and write down the one behavior that changed and the one that did not.
5. If the result differs from the release body, stop the rollout and attach the log, screenshot, or transcript to the version record.

openclaw --version
openclaw gateway status
openclaw doctor

Close the loop with evidence

A useful maintenance record names the changed behavior, the behavior that stayed stable, the identity that approved the test, and the step that would undo it. Preserve that record beside the release tag.

Edge cases in this tag

Installer. let the local-prefix CLI installer use Alpine's apk Node.js, npm, and Git packages on musl Linux instead of downloading glibc Node tarballs that fail node:sqlite. Record the failure that this correction addresses before declaring the upgrade complete.

Checks. prefilter tracked conflict-marker scans so changed checks avoid reading every repository file on clean runs. This is the edge case to include in a regression check, especially when the release is pre-release.

The authoritative record is the v2026.5.25-beta.1 GitHub release. The release index gives the surrounding sequence, while the getting-started guide supplies the operational background.