Repository navigation
chore(sparrow): release ag2-sparrow 0.6.1 - #5897
Conversation
The Hatch held-task line (#5828), Hatch self-upgrade to the served target_version (#5676), the served-setup fetch from the broker (#5855), served Hatch wording (#5665), seat-bridge env and token fixes (#5893, #5859), and the claude-local/codex-local connect commands (#5860, #5761). Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
|
The rerun's reds aren't from this one-line bump (Mars, Qingyun's Sutando seat): python standalone 1/9 is |
|
@cla-assistant check |
Coverage Gate✅ Diff coverage PASSES the 95% bar. Whole-tree (informational): 94%. Diff CoverageDiff: origin/main...HEAD, staged and unstaged changes
Summary
|
qingyun-wu
left a comment
There was a problem hiding this comment.
Approved at 0f334c2 (agent-assisted review by Mars, Qingyun's Sutando seat). This head is edf7774, which I reviewed, plus sonichi's merge of main. The diff against main is exactly the one-line ag2_sparrow.__version__ bump to 0.6.1 (dist version single-sourced from it, following sparrow-v0.6.0 and PyPI 0.6.0), unchanged, and no commit is under qingyun-wu. CI is fully green now that #5900 has landed (35 pass, 3 skipped).
|
@cla-assistant check |
What changed, and why
Releases ag2-sparrow 0.6.1. It's a one-line
__version__bump from0.6.1.dev1to0.6.1, the same shape as #5811 (0.6.0) and #5675 (0.5.4). The package keeps no CHANGELOG, and earlier bumps changed only__init__.py, so this one does too.The owner approved this release. Two things need it:
hold(fix(ag2-sparrow): a Hatch task parked for the owner's go posts one line and reports held, not stuck #5828). The backend's served Hatch setup list (on itsdevbranch) already requiresag2-sparrow>=0.6.1because itsworker_notetells workers to runhold.target_version. Getting from 0.6.0 to 0.6.1 is still a manual install.User-visible changes since 0.6.0
From
git log sparrow-v0.6.0..origin/main -- packages/ag2-sparrow:Hatch
held, not stuck (ag2-sparrow-hatch hold).target_version, with an offline rollback. Its limits:target_versionwith itstarget_sha256.stop-failed, as the README says.setup_revision, Sparrow fetches its served setup from the broker's/v1/setup.Connectors
ag2-sparrow-claude-local connect --base --code.ag2-sparrow-codex-localhosts the Codex connector under the shared supervised-session lifecycle #5761:ag2-sparrow-codex-localruns the Codex connector under the shared supervised-session lifecycle.Seat bridges and sparrowd
wire_bridgecomputes, not a vault token.Gateway and delivery
Other
Files to look at first
packages/ag2-sparrow/ag2_sparrow/__init__.py, the only change.What this proves
N/A. This is a version bump. The publish workflow refuses a tag that differs from the packaged
__version__, and refuses a version that already exists on PyPI.Feature/documentation consistency
N/A. No behavior changes here. The README already documents every feature listed above.
Before/after evidence
Tests
Run at head on macOS (python3 3.14):
python3 packages/ag2-sparrow/tests/test_hatch_served_setup.py: Ran 226 tests, OK. This suite holds the self-upgrade decision tests. They pininstalled=explicitly, so the bump doesn't change their outcome.python3 packages/ag2-sparrow/tests/test_hatch_adapter.py: Ran 415 tests, OK (skipped=2).bash scripts/review-checks.sh --diff <(git diff origin/main...HEAD): PASS.Live path
N/A. Nothing runs until the tag is pushed and a host installs 0.6.1.
Release steps, after merge
publish-sparrow.ymlthen publishes to PyPI over OIDC and attaches the.pyzand linux binaries to the GitHub Release. Tagging needs the owner's word, which has been given for this release.Deployment risks and follow-ups
package.target_versionandtarget_sha256arenullon both backend branches. Setting them is a backend change, made after PyPI publishes the 0.6.1 wheel.min_packagebump has to wait for this release. The backend change on itsdevbranch raisesmin_packageto>=0.6.1. It must not reachmainbefore 0.6.1 is on PyPI.🤖 Generated with Claude Code