You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Track the remaining runtime-boundary hardening work that was intentionally left after the narrower fixes in #412 and #413.
This is a tracker issue, not a directly implementable PR. It should stay open only while it helps keep the child tasks visible. Each implementation must happen through a smaller child issue or PR with its own scope and verification.
Scope
In scope for this tracker:
Keep the runtime-boundary follow-up map current.
Split broad architecture work into review-sized child issues before implementation.
Link child PRs or issues back here when they land.
Close this tracker only when the remaining useful follow-ups are either completed, superseded, or moved into clearer standalone issues.
Remaining work here should stay focused on runtime-boundary diagnostics that are still missing, especially route-action and interaction-lock-specific signals where needed.
Prefer data that helps debug packaged-app failures and agent-led triage.
Out of scope
Do not implement all four areas in one PR.
Do not rewrite the router or change route shapes from this tracker alone.
Do not replace Solid, Electron, the sidecar, or the SDK layer.
Do not include a normalized entity-store rewrite, timeline virtualization, or broad prefetch scheduler work.
Do not create local-only design docs unless they are tied to an implementation PR.
Add focused unit tests for the boundary contract being changed.
Add or update the smallest relevant user-path E2E coverage, especially delayed route stability and post-interaction clickability.
Run targeted app/ui typechecks and affected Playwright specs.
Confirm there are no route-shape, product-copy, or broad layout changes unless that child issue explicitly asks for them.
Priority
P2.
This is important architecture hardening, but concrete user-visible or release-blocking regressions should stay in narrower P1 issues. Do not raise this tracker just because one child task becomes urgent; raise that child issue instead.
Execution mode
Agent should propose a small child issue or PR plan before editing. Keep each implementation independently reviewable and reversible.
Goal
Track the remaining runtime-boundary hardening work that was intentionally left after the narrower fixes in #412 and #413.
This is a tracker issue, not a directly implementable PR. It should stay open only while it helps keep the child tasks visible. Each implementation must happen through a smaller child issue or PR with its own scope and verification.
Scope
In scope for this tracker:
Tracked boundary areas:
Navigation ownership
openProject(projectRoot),openSession(directory, sessionID), andnewSession(directory)landed in [Task] Harden interaction locks and session route ownership #413.Interaction locks
Desktop runtime manifest
Diagnostics
Out of scope
Relevant files or context
Related context:
Likely areas:
packages/app/src/pages/layout.tsxpackages/app/src/pages/layout/helpers.tspackages/app/src/pages/session/helpers.tspackages/ui/src/components/resize-handle.tsxpackages/desktop-electron/electron-builder.config.tspackages/desktop-electron/scripts/prepare-embedded-server.tspackages/app/e2e/sidebar/packages/app/e2e/commands/Verification
For each child PR or child issue:
Priority
P2.
This is important architecture hardening, but concrete user-visible or release-blocking regressions should stay in narrower P1 issues. Do not raise this tracker just because one child task becomes urgent; raise that child issue instead.
Execution mode
Agent should propose a small child issue or PR plan before editing. Keep each implementation independently reviewable and reversible.