~95% on SimpleQA (e.g. Qwen3.6-27B on a 3090). Supports all local and cloud LLMs (llama.cpp, Ollama, Google, ...). 10+ search engines - arXiv, PubMed, your private documents. Everything Local & Encrypted.
Python academiaanthropicarxivbravedeep-researchencryptionhome-automationhomeserverlocallocal-deep-researchlocal-llmmistralollamaopenaipubmedresearchresearch-toolretrieval-augmented-generationsearxngself-hosted
fix(ci): fail fast when the responsive-UI test server isn't ready (fixes ERR_CONNECTION_REFUSED) (#4779)
LearningCircuitfd09b40784b4
diff --git a/.github/workflows/responsive-ui-tests-enhanced.yml b/.github/workflows/responsive-ui-tests-enhanced.ymlindex 17a87d41a1..34fa0069bf 100644--- a/.github/workflows/responsive-ui-tests-enhanced.yml+++ b/.github/workflows/responsive-ui-tests-enhanced.yml@@ -146,22 +146,45 @@ jobs:echo "Server PID: $SERVER_PID"# Wait for server to start- for i in {1..30}; do- if curl -fsS --connect-timeout 2 --max-time 5 http://127.0.0.1:5000 2>/dev/null; then+ SERVER_READY=false+ for i in {1..60}; do+ # Probe the health endpoint, not bare `/`: `/` 302-redirects to+ # /auth/login and `curl -f` treats a 302 as success the instant+ # the socket binds — before DB/app init finishes — so it can pass+ # while the app is still not ready. /api/v1/health is a true+ # readiness signal and matches the docker-tests.yml gate.+ if curl -fsS --connect-timeout 2 --max-time 5 http://127.0.0.1:5000/api/v1/health 2>/dev/null; thenecho "Server is ready after $i seconds"+ SERVER_READY=truebreakfi# Check if process is still running- if ! kill -0 $SERVER_PID 2>/dev/null; then+ if ! kill -0 "$SERVER_PID" 2>/dev/null; thenecho "Server process died!"echo "Server log:"cat server.logexit 1fi- echo "Waiting for server... ($i/30)"+ echo "Waiting for server... ($i/60)"sleep 1done+ # Fail fast if the server is alive but never became ready in time.+ # Without this guard the loop fell through silently on a slow-but-+ # not-crashed start, so "Register CI test user" and the UI tests ran+ # against a server that was not yet listening and failed with a+ # confusing `net::ERR_CONNECTION_REFUSED` in the *test* step instead+ # of a clear startup failure here. Window was also bumped 30s -> 60s+ # to match the proven puppeteer-e2e-tests.yml / docker-tests.yml+ # startup gates (Postgres-backed boot can exceed 30s on a loaded+ # runner).+ if [ "$SERVER_READY" != "true" ]; then+ echo "❌ Server did not become ready within 60 seconds"+ echo "Server log:"+ cat server.log+ exit 1+ fi+- name: Register CI test userworking-directory: tests/ui_testsrun: node register_ci_user.js http://127.0.0.1:5000