test(editor): Add comprehensive instance AI e2e tests#28326

Draft

mutdmour wants to merge 19 commits intomasterfrom

feature/instance-ai-tabs

Commits on Apr 10, 2026

fix(editor): Polish instance AI artifact preview tabs (no-changelog)

mutdmour
and
claude
committed
fix exec finished bug
mutdmour
committed
test(editor): Add e2e test for instance AI preview execution state bug (no-changelog)

mutdmour
and
claude
committed
fix(editor): Relay all pending execution events before sending executionFinished (no-changelog)

mutdmour
and
claude
committed
test(editor): Add comprehensive instance AI e2e tests (no-changelog)

mutdmour
and
claude
committed
feat: Add trace replay infrastructure for instance AI e2e tests (no-changelog)

mutdmour
and
claude
committed
fix: Use system prompt body matcher for proxy expectations (no-changelog)

mutdmour
and
claude
committed
fix: Ensure recording mode when real API key is present (no-changelog)

mutdmour
and
claude
committed
chore: Re-record proxy expectations with body matchers (no-changelog)

mutdmour
and
claude
committed
fix: Set unlimited on last LLM expectation, not last file (no-changelog)

mutdmour
and
claude
committed
fix: Share TraceIndex across follow-up runs in trace replay (no-changelog)

mutdmour
and
claude
committed
fix: Wait for agent completion in sidebar tests to prevent flakiness (no-changelog)

mutdmour
and
claude
committed
fix: Remove contaminated proxy expectations and drain background tasks between tests (no-changelog)

mutdmour
and
claude
committed
fix: Retry proxy requests with exponential backoff to handle ECONNRESET (no-changelog)

mutdmour
and
claude
committed
fix: Use identity-based thread lookups to prevent cross-test contamination (no-changelog)

mutdmour
and
claude
committed
docs: Add detailed documentation for Instance AI e2e test infrastructure (no-changelog)

mutdmour
and
claude
committed
refactor: Add unit tests, extract test controller, and remove irrelevant proxy expectations (no-changelog)

mutdmour
and
claude
committed
docs: Clarify robustness guarantees and re-recording triggers for Instance AI e2e tests (no-changelog)

mutdmour
and
claude
committed
docs: Move robustness section after re-recording and fix inaccurate claims (no-changelog)

mutdmour
and
claude
committed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

test(editor): Add comprehensive instance AI e2e tests#28326

test(editor): Add comprehensive instance AI e2e tests#28326
mutdmour wants to merge 19 commits intomasterfrom
feature/instance-ai-tabs

Commits on Apr 10, 2026

fix(editor): Polish instance AI artifact preview tabs (no-changelog)

fix exec finished bug

test(editor): Add e2e test for instance AI preview execution state bug (no-changelog)

fix(editor): Relay all pending execution events before sending executionFinished (no-changelog)

test(editor): Add comprehensive instance AI e2e tests (no-changelog)

feat: Add trace replay infrastructure for instance AI e2e tests (no-changelog)

fix: Use system prompt body matcher for proxy expectations (no-changelog)

fix: Ensure recording mode when real API key is present (no-changelog)

chore: Re-record proxy expectations with body matchers (no-changelog)

fix: Set unlimited on last LLM expectation, not last file (no-changelog)

fix: Share TraceIndex across follow-up runs in trace replay (no-changelog)

fix: Wait for agent completion in sidebar tests to prevent flakiness (no-changelog)

fix: Remove contaminated proxy expectations and drain background tasks between tests (no-changelog)

fix: Retry proxy requests with exponential backoff to handle ECONNRESET (no-changelog)

fix: Use identity-based thread lookups to prevent cross-test contamination (no-changelog)

docs: Add detailed documentation for Instance AI e2e test infrastructure (no-changelog)

refactor: Add unit tests, extract test controller, and remove irrelevant proxy expectations (no-changelog)

docs: Clarify robustness guarantees and re-recording triggers for Instance AI e2e tests (no-changelog)

docs: Move robustness section after re-recording and fix inaccurate claims (no-changelog)