chore(release): v0.5.44 - #134
Merged
Merged
Conversation
Ships the client failure diagnosis merged in cc852a6 (#133). Until this release the fix is on main and nowhere else, so every installed CLI and the broker on the fleet still collapse DNS failure, connection refused, connect timeout, read timeout and TLS failure into a single "unreachable" string. Those cases have opposite remedies: refused means the path is fine and the broker is not running; a dropped packet means the broker is probably healthy and a firewall or tailnet ACL is in the way. On 2026-09-02 that ambiguity was read the wrong way round against a broker that was up and serving, and the error boundary's standing advice — "is the broker running?" — pointed at the one layer that was not the problem. The MCP half matters more than the CLI half. An agent cannot ssh to the host to check, so from a bare "timed out" its obvious inference is that the broker is down, and acting on that means restarting a service that is fine. Agents now receive the kind and the hint, and the call log records the specific transport_<kind> rather than a flat transport. No migration, no config change, no operator action.
musharna
marked this pull request as ready for review
September 3, 2026 03:31
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Ships the client failure diagnosis merged in
cc852a6(#133).Until this release the fix is on
mainand nowhere else — every installed CLI and the broker on the fleet still collapse DNS failure, connection refused, connect timeout, read timeout and TLS failure into oneunreachablestring. Those cases have opposite remedies, and on 2026-09-02 the ambiguity was read the wrong way round against a broker that was up and serving.Contents: one changelog fragment (
broker-failure-diagnosis.fixed.md), rolled into[0.5.44].Pre-release checks, run locally:
ruff check+ruff format --check— cleanmypy src/jobd src/job_cli— no issues in 48 source filespytest -m "not live"— 1352 passed, 9 deselectedscripts/quickstart-dry-run.sh dist/jobd-0.5.44-py3-none-any.whl— all steps passed, and step 1 reportedversion: 0.5.44, confirming it exercised this wheel rather than falling back to PyPI's 0.5.43No migration, no config change, no operator action.