testIntegration > Autocomplete.testAutocomplete fails at Autocomplete.kt:42 (latent bug masked by FFmpeg 7.1 404)
Maintainers usually reply within 3 days
Nobody has claimed this yet.
Assessment
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Newbie friendliness
- 48/100
- Issue type
- Bug
- Clarity
- Mostly clear
- Activity status
- Active
- Tech stack
- github-actions, kotlin
- Domain
- ci-cd, testing-qa
Research direction
Start with extensions/intellij/src/testIntegration/kotlin/com/github/continuedev/continueintellijextension/Autocomplete.kt and its testAutocomplete() assertion, then run testIntegration with the Setup FFmpeg step passing. Inspect the TestLLM fixture at extensions/intellij/src/testIntegration/kotlin/com/github/continuedev/continueintellijextension/test-continue/config.json, build.gradle.kts, and the jb-failure-report artifact to determine why the expected response is absent. Done means the test reliably passes in CI or the issue's chosen temporary resolution is documented.
Written by the indexing model from the issue text.
Description
Summary
:testIntegration > Autocomplete.testAutocomplete() fails at Autocomplete.kt:42 on every PR that successfully passes the Setup FFmpeg step. The test has been latent for ~11 months — it was added on 2025-09-19 via PR #7690 ("test: use intellij-ide-starter for e2e tests", CON-3723) but has effectively never run to completion in CI because the AnimMouse/setup-ffmpeg@v1 version: 7.1 pin in .github/actions/run-jetbrains-tests/action.yml returns 404 first (issue #13164).
Failure detail
- Test class:
com.github.continuedev.continueintellijextension.Autocomplete - Test method:
testAutocomplete() - Failing assertion (Autocomplete.kt:42):
assertTrue(text.contains("TEST_LLM_RESPONSE_0")) - Error:
expected: <true> but was: <false>(org.opentest4j.AssertionFailedError) - Duration to failure: 1m04s (the test runs to completion; the assertion fails at the end)
- IDE under test:
IC-243.21565.193(IntelliJ Community 2024.3) - Test framework:
intellij-ide-starter(TestFrameworkType.Starter), pinned inextensions/intellij/build.gradle.kts - JUnit: Jupiter 5.7.1
- Artifact (video + test reports):
jb-failure-reportfrom run 32468648181 — video atvideo/testAutocomplete_recording_2026_21_08_09_43_36.avi
Why this surfaces now
Issue #13164 reported that the FFmpeg 7.1 pin returns 404. PR #13169 fixes it by pinning to 8.1 + adding a preflight check. Once the FFmpeg step passes on the PR head (2e6fadcbd), testIntegration runs and this assertion fires.
Why this is pre-existing (not caused by the FFmpeg fix)
- The PR diff is only
.github/actions/run-jetbrains-tests/action.yml(parameterized version + preflight check). No JetBrains plugin source is touched. - The test framework, IntelliJ version, and
TestLLMfixture have not changed since 2025-09-19. - The test was effectively dead code in CI for the entire window between #7690 and #13164: main-branch
version: 7.1always 404s the FFmpeg step first, sotestIntegrationis never reached. - The closest prior PR (#13165 by ScrewTSW) was self-closed on 2026-08-19 before CI ran.
Likely root causes (not exhaustive)
- Autocomplete never fires in the new IDE under test. The
tab()key afterenterText("TEST_USER_MESSAGE_0")+space()may not trigger Continue's inline completion — could be a default-settings issue (autocomplete disabled) or a keystroke routing issue. - TestLLM not loaded — the test fixture config in
extensions/intellij/src/testIntegration/kotlin/com/github/continuedev/continueintellijextension/test-continue/config.json(inferred fromCONTINUE_GLOBAL_DIRinbuild.gradle.kts) may not be picked up by the 2024.3 IDE under test. - Prompt-to-response mismatch —
TestLLM.findResponse()requires the regex[^0-9]*TEST_USER_MESSAGE_(\d+)[^0-9]*. If the IDE injects surrounding text that breaks the digit boundaries,findResponsereturnsundefinedand the stream yieldsPROMPT: <full prompt>instead ofTEST_LLM_RESPONSE_0. - Timing — the test waits
2.secondsafter typing, which may not be enough for the autocomplete provider to compute a response under the new test framework.
Suggested next step
Pick one:
- (A) Investigate and fix the test (preferred — the test should be a working CI gate).
- (B) Mark the test
@Disabledwith a TODO referencing this issue until a fix lands. - (C) Accept that this test is flaky in the new framework and remove it.
Related
- Fix PR (unblocks this test from being reached): #13169 "fix(ci): parameterize pinned FFmpeg version with explicit preflight check (BtbN dropped 7.1)"
- Root-cause bug: #13164
- Original test introduction: #7690 (CON-3723), commit
507cdfc61aon 2025-09-19 - Failed CI run with artifacts: https://github.com/continuedev/continue/actions/runs/32468648181
- Dominant language
- TypeScript
- Stars
- 36.1k
- Forks
- 5.4k
- PR merge metrics
- No merged PRs in 30d
Getting set up
- No Dockerfile or Docker Compose file
- Has a pull request template
- Read the contributing guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
More from continuedev/continue
-
Difficulty 2/5 1-3 hours Newbie friendliness 72/100
continuedev/continue#13291 ·
Maintainers usually reply within 3 days
-
Ollama provider drops `tools` after the first tool call, breaking multi-step agent loopsPossibly taken @dvd233 claimed this 23 days ago. Open
Difficulty 2/5 1-3 hours Newbie friendliness 78/100
continuedev/continue#13254 · 2 comments ·
Maintainers usually reply within 3 days
-
[Bug] LanceDbIndex interpolates paths into LanceDB SQL predicates — apostrophes in file/workspace paths break index deletes (stale chunks stay retrievable) and retrievalPossibly taken @seethinajayadileep claimed this 29 days ago. Open
Difficulty 2/5 1-3 hours Newbie friendliness 82/100
continuedev/continue#13233 ·
Maintainers usually reply within 3 days
-
docs: PZERO OpenAI-compatible apiBase examplePossibly taken @dajiaohuang claimed this 32 days ago. Open
Difficulty 1/5 Under an hour Newbie friendliness 85/100
continuedev/continue#13198 · 1 comment ·
Maintainers usually reply within 3 days
-
Difficulty 1/5 Under an hour Newbie friendliness 88/100
continuedev/continue#13197 ·
Maintainers usually reply within 3 days
All issues in continuedev/continue
Similar issues
-
Difficulty 2/5 1-3 hours Newbie friendliness 72/100
Maintainers usually reply within 1 day
-
Difficulty 2/5 1-3 hours Newbie friendliness 68/100
Maintainers usually reply within 1 day
-
Difficulty 2/5 1-3 hours Newbie friendliness 68/100
-
Difficulty 2/5 1-3 hours Newbie friendliness 72/100
Maintainers usually reply within 4 days
-
Difficulty 2/5 1-3 hours Newbie friendliness 78/100
Maintainers usually reply within 1 day