* fix(macos): workaround App Store Connect dSYM validation bug for App.framework
* trigger ci
* test: mock device and package plugins and increase timeout
- Increase sync loop timeout in thread workspace binding test to avoid flakiness
- Mock device_info and package_info plugins for gateway runtime tests
- Update pubspec.yaml version
* test: fix missing plugin in runtime_controllers_settings_account_test
* build: make sync-version.sh auto-increment build number
---------
Co-authored-by: Haitao Pan <manbuzhe2009@qq.com>
* ci: gate TestFlight behind opt-in toggle + Xcode 27 build fixes
TestFlight is now opt-in (default OFF). A workflow_dispatch boolean
`enable_testflight` (or the `ENABLE_TESTFLIGHT` repo variable) drives a
`prepare.outputs.testflight_enabled` flag that gates the macOS
app-store-pkg build leg and both testflight_ios/testflight_macos upload
legs. Missing Apple signing secrets no longer fail the normal DMG/IPA
release path (package-macos-app-store-pkg.sh hard-exits without them).
Xcode 27 build compatibility:
- Align Apple deployment targets so no pod sits below the app minimum
(Xcode 27 rejects this): macOS pods + RunnerTests -> 15.6, iOS pods
-> 15.5 to match the Runner targets.
- Add a `lipo` shim (scripts/xcode-tools/lipo) wired onto PATH in the
iOS/macOS build phases; Xcode 27 only accepts one `-verify_arch`
architecture per call while Flutter passes them all at once.
- macOS project hygiene: correct PrivacyInfo.xcprivacy path, set app
display name + LSApplicationCategoryType.
* test: make temp-dir cleanup resilient to concurrent-write races
The assistant execution target tests deleted their temp HOME/workspace
dirs with a raw recursive delete in addTearDown. A background flush
(e.g. controller dispose still persisting state) can keep writing into
the dir while the delete walks it, so the delete races and fails with
"Directory not empty" (errno 39), failing the test on CI.
Route all unguarded teardown deletes through the existing
_resilientDelete helper (re-check existence + retry), and harden that
helper so its final fallback never re-throws — a temp-dir cleanup
failure must never fail a test.
---------
Co-authored-by: Haitao Pan <manbuzhe2009@qq.com>
Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>
Every gateway turn's prompt prefix injected three near-duplicate absolute paths:
currentTaskWorkspace + localWorkspace + remoteWorkspaceHint. localWorkspace is the
App's LOCAL thread dir (~/.xworkmate/threads/...) which the gateway agent cannot
access, and remoteWorkspaceHint duplicates currentTaskWorkspace. The conflicting
paths leave the agent unsure where to work and can block conversation continuation.
For gateway turns the prompt now carries only currentTaskWorkspace (the plugin owns
the artifact scope); localWorkspace is kept only for non-gateway (local agent runs
there); remoteWorkspaceHint is dropped when equal to currentTaskWorkspace. sessionKey
is kept (short, not a path). UI is unaffected (chat bubble shows the raw user message;
the prompt-debug parser only special-cases Execution context / Preferred skills /
Attached files). Tests updated; assistant_execution_target_test green (74).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
- Remove the unexpected property at the root of gateway task metadata.
Keep the value nested in xworkmateTaskArtifactContract where the
OpenClaw chat.send schema allows it (-32002: invalid chat.send params).
- Drop dead local vars and the unused asInt helper in OpenClaw task
association parsing.
- Remove the obsolete 'sendChatMessage restarts before handling
OpenClaw artifact guard results' test superseded by the new terminal
artifact failure test.
