feat: Fast ci virtual cycle

.github/workflows/dev_cycle_fast.yml

@@ -0,0 +1,42 @@
+name: CI Fast Dev Cycle
+
+on:
+  pull_request:
+  workflow_dispatch:
+
+jobs:
+  lint_and_smoke:
+    runs-on: ubuntu-latest
+    timeout-minutes: 35
+
+    steps:
+      - name: Set up Rust toolchain
+        uses: hecrj/setup-rust-action@v2
+        with:
+          rust-version: stable
+
+      - name: Check out the code
+        uses: actions/checkout@v4
+
+      - name: Install Clippy
+        run: rustup component add clippy
+
+      - name: Run Clippy
+        run: cargo clippy --all-targets --all-features -- -D warnings
+
+      - name: Install cargo-nextest
+        run: cargo install cargo-nextest --locked
+
+      - name: Run smoke tests (fast profile)
+        env:
+          RUST_MIN_STACK: 8388608
+        run: >
+          cargo nextest run
+          --profile dev-cycle
+          --status-level=all
+          --no-fail-fast
+          -E 'test(known_group_persistence)
+          | test(group_name_persistence)
+          | test(repo_persistence)
+          | test(test_rpc_service_init)
+          | test(test_rpc_client)'

nextest.toml

@@ -2,11 +2,23 @@
 # Veilid network tests can take several minutes due to DHT operations and network setup
 
 [profile.default]
-# Set timeout to 10 minutes per test (Veilid network operations can be slow)
-test-timeout = "600s"
+# Set timeout budget for slow Veilid tests
+slow-timeout = { period = "60s", terminate-after = 10 }
 
 # Retry flaky tests up to 3 times
 retries = 3
 
 # Run tests serially to avoid Veilid network conflicts
 test-threads = 1
+
+# Fast feedback profile for development cycles.
+# This profile intentionally keeps a tighter timeout/retry budget.
+[profile.dev-cycle]
+slow-timeout = { period = "30s", terminate-after = 4 }
+retries = 1
+test-threads = 1
+
+[[profile.dev-cycle.overrides]]
+filter = 'test(repo_persistence) | test(test_auto_create_repo_on_join_group)'
+slow-timeout = { period = "45s", terminate-after = 4 }
+retries = 2