ci(065): fix D1 retrieval job — create data_dir + single-step server lifecycle

The retrieval-d1 job failed: `mcpproxy serve` exited immediately with
"data_dir: directory does not exist" (serve refuses to create a missing
data_dir), and the server was backgrounded in a separate step from the readiness
poll (a process backgrounded in one step is reaped when that step's shell exits).

Fix: mkdir -p the data_dir, and boot + readiness-poll + run the scorer in ONE
step (shared shell) with a trap that stops the server however the step ends; also
fail fast if the server process dies during startup. D2 gate unaffected.

Verified: mcpproxy boots and serves /api/v1/status locally with the created
data_dir; actionlint clean.

Related #555 datasets; MCP-742.

Co-Authored-By: Paperclip <noreply@paperclip.ing>

Author: Dumbris

.github/workflows/eval.yml

@@ -131,51 +131,57 @@ jobs:
       - name: Build mcpproxy
         run: go build -o mcpproxy ./cmd/mcpproxy
 
-      - name: Start mcpproxy (7 reference servers)
+      # Boot, poll and score in ONE step: a server backgrounded in a separate
+      # step is reaped when that step's shell exits, so start + wait + run must
+      # share a shell. The trap stops the server however this step ends.
+      - name: Run D1 retrieval gate (boot mcpproxy + score)
+        working-directory: ${{ github.workspace }}
+        env:
+          DS: ${{ github.workspace }}/specs/065-evaluation-foundation/datasets
         run: |
+          set -uo pipefail
+          base="http://127.0.0.1:8092"; key="eval-corpus-snapshot"
+          # data_dir must exist — `serve` refuses to create a missing one.
+          mkdir -p "$RUNNER_TEMP/eval"
           ./mcpproxy serve \
-            --config specs/065-evaluation-foundation/datasets/snapshot-servers.config.json \
+            --config "$DS/snapshot-servers.config.json" \
             --data-dir "$RUNNER_TEMP/eval" \
             --listen 127.0.0.1:8092 \
             --log-level info > "$RUNNER_TEMP/mcpproxy.log" 2>&1 &
-          echo "MCPPROXY_PID=$!" >> "$GITHUB_ENV"
+          server_pid=$!
+          trap 'kill "$server_pid" 2>/dev/null || true' EXIT
 
-      - name: Wait for index readiness
-        run: |
-          base="http://127.0.0.1:8092"; key="eval-corpus-snapshot"
+          ready=0
           for i in $(seq 1 60); do
+            if ! kill -0 "$server_pid" 2>/dev/null; then
+              echo "::error::mcpproxy process exited during startup"
+              break
+            fi
             if curl -fsS -H "X-API-Key: $key" "$base/api/v1/status" >/dev/null 2>&1; then
               n="$(curl -fsS -H "X-API-Key: $key" "$base/api/v1/index/search?q=file&limit=5" \
                    | python3 -c 'import sys,json;d=json.load(sys.stdin);print(len(d.get("tools",d.get("results",[]))))' 2>/dev/null || echo 0)"
               echo "attempt $i: index search returned $n result(s)"
-              [ "$n" -ge 1 ] && { echo "Index ready."; exit 0; }
+              [ "$n" -ge 1 ] && { ready=1; echo "Index ready."; break; }
             else
               echo "attempt $i: server not up yet"
             fi
             sleep 5
           done
-          echo "::error::mcpproxy index did not become ready in time"
-          tail -50 "$RUNNER_TEMP/mcpproxy.log" || true
-          exit 1
+          if [ "$ready" != 1 ]; then
+            echo "::error::mcpproxy index did not become ready in time"
+            echo "----- mcpproxy.log (tail) -----"; tail -80 "$RUNNER_TEMP/mcpproxy.log" || true
+            exit 1
+          fi
 
-      - name: Run D1 retrieval gate
-        working-directory: mcp-eval
-        env:
-          DS: ${{ github.workspace }}/specs/065-evaluation-foundation/datasets
-        run: |
-          PYTHONPATH=src uv run python -m mcp_eval.cli retrieval \
+          ( cd "$GITHUB_WORKSPACE/mcp-eval" && PYTHONPATH=src uv run python -m mcp_eval.cli retrieval \
             --corpus "$DS/corpus_v1.tools.json" \
             --golden "$DS/retrieval_golden_v1.json" \
             --baseline "$DS/baseline_v1.json" \
             --tolerance 0.05 \
             --runs 1 \
-            --base-url http://127.0.0.1:8092 \
-            --api-key eval-corpus-snapshot \
-            --out-dir "${{ github.workspace }}/reports/retrieval"
-
-      - name: Stop mcpproxy
-        if: always()
-        run: kill "${MCPPROXY_PID}" 2>/dev/null || true
+            --base-url "$base" \
+            --api-key "$key" \
+            --out-dir "$GITHUB_WORKSPACE/reports/retrieval" )
 
       - name: Upload D1 reports
         if: always()