From 338ed770917e7159a2908c94c0016a46bc63496a Mon Sep 17 00:00:00 2001
From: PGijsbers <p.gijsbers@tue.nl>
Date: Wed, 7 Oct 2026 15:02:14 +0200
Subject: [PATCH 01/12] Add automatic test for dataset upload and processing

---
 scripts/data/test.arff               | 14 ++++++++++++
 scripts/evaluation_engine_dataset.sh | 10 ++++++++
 scripts/upload_dataset.sh            | 34 ++++++++++++++++++++++++++++
 scripts/wait_for_dataset.sh          | 27 ++++++++++++++++++++++
 4 files changed, 85 insertions(+)
 create mode 100644 scripts/data/test.arff
 create mode 100644 scripts/evaluation_engine_dataset.sh
 create mode 100755 scripts/upload_dataset.sh
 create mode 100755 scripts/wait_for_dataset.sh

diff --git a/scripts/data/test.arff b/scripts/data/test.arff
new file mode 100644
index 0000000..51ab11c
--- /dev/null
+++ b/scripts/data/test.arff
@@ -0,0 +1,14 @@
+@RELATION test-dataset
+@ATTRIBUTE foo NUMERIC
+@ATTRIBUTE bar NUMERIC
+@ATTRIBUTE bazz NUMERIC
+@ATTRIBUTE class {'a', 'b'}
+@DATA
+0,0,0,'a'
+0,0,1,'a'
+0,1,0,'a'
+0,1,1,'a'
+1,0,0,'b'
+1,0,1,'b'
+1,1,0,'b'
+1,1,1,'b'
diff --git a/scripts/evaluation_engine_dataset.sh b/scripts/evaluation_engine_dataset.sh
new file mode 100644
index 0000000..dd4f2ee
--- /dev/null
+++ b/scripts/evaluation_engine_dataset.sh
@@ -0,0 +1,10 @@
+#!/bin/bash
+# Tests the dataset upload and processing of the REST API and Evaluation Engine
+set -euo pipefail
+
+SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
+
+DATASET_ID=$("${SCRIPT_DIR}/upload_dataset.sh")
+echo "Uploaded dataset with ID: $DATASET_ID"
+
+"${SCRIPT_DIR}/wait_for_dataset.sh" "$DATASET_ID"
diff --git a/scripts/upload_dataset.sh b/scripts/upload_dataset.sh
new file mode 100755
index 0000000..97c767c
--- /dev/null
+++ b/scripts/upload_dataset.sh
@@ -0,0 +1,34 @@
+#!/bin/bash
+# Uploads a dataset to the REST API and prints the dataset ID to stdout.
+set -euo pipefail
+
+XML_DESCRIPTION='<oml:data_set_description xmlns:oml="http://openml.org/openml">
+	<oml:name>test_dataset</oml:name>
+	<oml:description>test</oml:description>
+	<oml:format>arff</oml:format>
+	<oml:creator>I</oml:creator>
+	<oml:collection_date>now</oml:collection_date>
+	<oml:language>en</oml:language>
+	<oml:licence>BSD (from scikit-learn)</oml:licence>
+	<oml:default_target_attribute>class</oml:default_target_attribute>
+	<oml:version_label>test</oml:version_label>
+	<oml:citation>citation</oml:citation>
+	<oml:original_data_url>https://www4.stat.ncsu.edu/~boos/var.select/diabetes.html</oml:original_data_url>
+	<oml:paper_url>url</oml:paper_url>
+</oml:data_set_description>'
+
+SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
+
+RESPONSE=$(curl -s -X POST "http://localhost:8000/api/v1/xml/data?api_key=normaluser" \
+  -F "description=<-;type=text/xml" <<< "${XML_DESCRIPTION}" \
+  -F "dataset=@${SCRIPT_DIR}/data/test.arff")
+
+DATASET_ID=$(echo "$RESPONSE" | sed -n 's/.*<oml:id>\([0-9]*\)<\/oml:id>.*/\1/p')
+
+if [ -z "$DATASET_ID" ]; then
+  echo "ERROR: Failed to parse dataset ID from response:" >&2
+  echo "$RESPONSE" >&2
+  exit 1
+fi
+
+echo "$DATASET_ID"
diff --git a/scripts/wait_for_dataset.sh b/scripts/wait_for_dataset.sh
new file mode 100755
index 0000000..6bdd0a9
--- /dev/null
+++ b/scripts/wait_for_dataset.sh
@@ -0,0 +1,27 @@
+#!/bin/bash
+# Waits for a dataset to become active.
+# Usage: wait_for_dataset.sh DATASET_ID [TIMEOUT]
+set -euo pipefail
+
+DATASET_ID="${1:?Usage: wait_for_dataset.sh DATASET_ID [TIMEOUT]}"
+TIMEOUT="${2:-120}"
+INTERVAL=5
+ELAPSED=0
+
+while [ "$ELAPSED" -lt "$TIMEOUT" ]; do
+  STATUS_RESPONSE=$(curl -s "http://localhost:8000/api/v1/xml/data/${DATASET_ID}")
+  STATUS=$(echo "$STATUS_RESPONSE" | sed -n 's/.*<oml:status>\([a-zA-Z_]*\)<\/oml:status>.*/\1/p')
+
+  echo "Status after ${ELAPSED}s: ${STATUS:-unknown}"
+
+  if [ "$STATUS" = "active" ]; then
+    echo "Dataset $DATASET_ID is active."
+    exit 0
+  fi
+
+  sleep "$INTERVAL"
+  ELAPSED=$((ELAPSED + INTERVAL))
+done
+
+echo "ERROR: Dataset $DATASET_ID did not become active within ${TIMEOUT} seconds."
+exit 1

From 40a65026d6614315719a7b05a7e30c000045caea Mon Sep 17 00:00:00 2001
From: PGijsbers <p.gijsbers@tue.nl>
Date: Wed, 7 Oct 2026 15:20:02 +0200
Subject: [PATCH 02/12] Add scripts to test creating tasks and task splits

---
 scripts/create_task.sh    | 33 +++++++++++++++++++++++++++++++++
 scripts/upload_dataset.sh |  6 +++++-
 scripts/wait_for_task.sh  | 31 +++++++++++++++++++++++++++++++
 3 files changed, 69 insertions(+), 1 deletion(-)
 create mode 100755 scripts/create_task.sh
 create mode 100755 scripts/wait_for_task.sh

diff --git a/scripts/create_task.sh b/scripts/create_task.sh
new file mode 100755
index 0000000..d09bc99
--- /dev/null
+++ b/scripts/create_task.sh
@@ -0,0 +1,33 @@
+#!/bin/bash
+# Creates a task for a given dataset and estimation procedure, prints the task ID to stdout.
+# Usage: create_task.sh DATASET_ID ESTIMATION_PROCEDURE_ID [TARGET_FEATURE]
+set -euo pipefail
+
+DATASET_ID="${1:?Usage: create_task.sh DATASET_ID ESTIMATION_PROCEDURE_ID [TARGET_FEATURE]}"
+ESTIMATION_PROCEDURE_ID="${2:?Usage: create_task.sh DATASET_ID ESTIMATION_PROCEDURE_ID [TARGET_FEATURE]}"
+TARGET_FEATURE="${3:-class}"
+
+TMPFILE=$(mktemp /tmp/task_desc.XXXXXX.xml)
+trap 'rm -f "$TMPFILE"' EXIT
+
+cat > "$TMPFILE" << EOF
+<oml:task_inputs xmlns:oml="http://openml.org/openml">
+  <oml:task_type_id>1</oml:task_type_id>
+  <oml:input name="source_data">${DATASET_ID}</oml:input>
+  <oml:input name="estimation_procedure">${ESTIMATION_PROCEDURE_ID}</oml:input>
+  <oml:input name="target_feature">${TARGET_FEATURE}</oml:input>
+</oml:task_inputs>
+EOF
+
+RESPONSE=$(curl -s -X POST "http://localhost:8000/api/v1/xml/task?api_key=normaluser" \
+  -F "description=@${TMPFILE};type=text/xml")
+
+TASK_ID=$(echo "$RESPONSE" | sed -n 's/.*<oml:id>\([0-9]*\)<\/oml:id>.*/\1/p')
+
+if [ -z "$TASK_ID" ]; then
+  echo "Server response:" >&2
+  echo "$RESPONSE" >&2
+  exit 1
+fi
+
+echo "$TASK_ID"
diff --git a/scripts/upload_dataset.sh b/scripts/upload_dataset.sh
index 97c767c..e031ad6 100755
--- a/scripts/upload_dataset.sh
+++ b/scripts/upload_dataset.sh
@@ -19,8 +19,12 @@ XML_DESCRIPTION='<oml:data_set_description xmlns:oml="http://openml.org/openml">
 
 SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
 
+TMPFILE=$(mktemp /tmp/dataset_desc.XXXXXX.xml)
+trap 'rm -f "$TMPFILE"' EXIT
+echo "$XML_DESCRIPTION" > "$TMPFILE"
+
 RESPONSE=$(curl -s -X POST "http://localhost:8000/api/v1/xml/data?api_key=normaluser" \
-  -F "description=<-;type=text/xml" <<< "${XML_DESCRIPTION}" \
+  -F "description=@${TMPFILE};type=text/xml" \
   -F "dataset=@${SCRIPT_DIR}/data/test.arff")
 
 DATASET_ID=$(echo "$RESPONSE" | sed -n 's/.*<oml:id>\([0-9]*\)<\/oml:id>.*/\1/p')
diff --git a/scripts/wait_for_task.sh b/scripts/wait_for_task.sh
new file mode 100755
index 0000000..4f9d3e6
--- /dev/null
+++ b/scripts/wait_for_task.sh
@@ -0,0 +1,31 @@
+#!/bin/bash
+# Waits until a task has associated split files.
+# Usage: wait_for_task.sh TASK_ID [TIMEOUT]
+set -euo pipefail
+
+TASK_ID="${1:?Usage: wait_for_task.sh TASK_ID [TIMEOUT]}"
+TIMEOUT="${2:-120}"
+INTERVAL=5
+ELAPSED=0
+
+while [ "$ELAPSED" -lt "$TIMEOUT" ]; do
+  RESPONSE=$(curl -s "http://localhost:8000/api/v1/xml/task/${TASK_ID}")
+  SPLITS_URL=$(echo "$RESPONSE" | sed -n 's/.*<oml:data_splits_url>\(.*\)<\/oml:data_splits_url>.*/\1/p')
+
+  if [ -n "$SPLITS_URL" ]; then
+    STATUS_CODE=$(curl -s -o /dev/null -w "%{http_code}" "$SPLITS_URL")
+    if [ "$STATUS_CODE" = "200" ]; then
+      echo "Task $TASK_ID has split files available."
+      exit 0
+    fi
+    echo "Splits URL exists but returned HTTP $STATUS_CODE after ${ELAPSED}s"
+  else
+    echo "No splits URL yet after ${ELAPSED}s"
+  fi
+
+  sleep "$INTERVAL"
+  ELAPSED=$((ELAPSED + INTERVAL))
+done
+
+echo "ERROR: Task $TASK_ID did not get split files within ${TIMEOUT} seconds."
+exit 1

From deb46c0b48acc7f114c55a698e23b43785290b9b Mon Sep 17 00:00:00 2001
From: PGijsbers <p.gijsbers@tue.nl>
Date: Thu, 8 Oct 2026 11:49:29 +0200
Subject: [PATCH 03/12] Add CI tests for dataset and task processing

---
 .github/actions/start-services/action.yaml | 34 ++++++++++++++
 .github/workflows/integration-test.yaml    | 53 ++++++++++++++++++++++
 2 files changed, 87 insertions(+)
 create mode 100644 .github/actions/start-services/action.yaml
 create mode 100644 .github/workflows/integration-test.yaml

diff --git a/.github/actions/start-services/action.yaml b/.github/actions/start-services/action.yaml
new file mode 100644
index 0000000..26572b0
--- /dev/null
+++ b/.github/actions/start-services/action.yaml
@@ -0,0 +1,34 @@
+name: Start Services
+description: Checkout the repo and start Docker Compose services
+
+inputs:
+  profiles:
+    description: "Comma-separated list of profiles to start (e.g. 'all' or 'rest-api,minio')"
+    required: false
+    default: "all"
+  services:
+    description: "Comma-separated list of specific services to start (e.g. 'database,elasticsearch')"
+    required: false
+    default: ""
+
+runs:
+  using: composite
+  steps:
+    - name: Checkout
+      uses: actions/checkout@v4
+
+    - name: Start services
+      shell: bash
+      run: |
+        PROFILE_FLAGS=""
+        for profile in $(echo "${{ inputs.profiles }}" | tr ',' ' '); do
+          PROFILE_FLAGS="$PROFILE_FLAGS --profile $profile"
+        done
+        SERVICES=$(echo "${{ inputs.services }}" | tr ',' ' ')
+        docker compose $PROFILE_FLAGS up -d --build --wait $SERVICES
+      timeout-minutes: 10
+
+    - name: Show service status
+      if: always()
+      shell: bash
+      run: docker compose ps
diff --git a/.github/workflows/integration-test.yaml b/.github/workflows/integration-test.yaml
new file mode 100644
index 0000000..2dc8364
--- /dev/null
+++ b/.github/workflows/integration-test.yaml
@@ -0,0 +1,53 @@
+name: Integration Test
+
+on:
+  pull_request:
+    paths:
+      - "docker-compose.yaml"
+      - "scripts/**"
+      - "config/**"
+
+jobs:
+  dataset-processing:
+    runs-on: ubuntu-latest
+    timeout-minutes: 15
+
+    steps:
+      - name: Start services
+        uses: ./.github/actions/start-services
+        with:
+          profiles: "rest-api,evaluation-engine"
+
+      - name: Create dataset
+        id: dataset
+        run: |
+          DATASET_ID=$(bash scripts/upload_dataset.sh)
+          echo "Dataset ID: $DATASET_ID"
+          echo "id=$DATASET_ID" >> "$GITHUB_OUTPUT"
+
+      - name: Dataset Processing
+        run: bash scripts/wait_for_dataset.sh "${{ steps.dataset.outputs.id }}" 180
+
+  task-processing:
+    runs-on: ubuntu-latest
+    timeout-minutes: 15
+
+    steps:
+      - name: Start services
+        uses: ./.github/actions/start-services
+        with:
+          profiles: "rest-api,evaluation-engine"
+
+      - name: Create task
+        id: task
+        run: |
+          TASK_ID=$(bash scripts/create_task.sh 129 1)
+          echo "Task ID: $TASK_ID"
+          echo "id=$TASK_ID" >> "$GITHUB_OUTPUT"
+
+      - name: Task Processing
+        run: bash scripts/wait_for_task.sh "${{ steps.task.outputs.id }}" 180
+
+      - name: Show logs on failure
+        if: failure()
+        run: docker compose logs --tail=50

From 0f78f0a6ce030bd19dacbccf0908e43064593d30 Mon Sep 17 00:00:00 2001
From: PGijsbers <p.gijsbers@tue.nl>
Date: Thu, 8 Oct 2026 11:56:45 +0200
Subject: [PATCH 04/12] Move checkout outside of action, otherwise it can't be
 found

---
 .github/actions/start-services/action.yaml |  5 +----
 .github/workflows/integration-test.yaml    | 10 ++++++++++
 2 files changed, 11 insertions(+), 4 deletions(-)

diff --git a/.github/actions/start-services/action.yaml b/.github/actions/start-services/action.yaml
index 26572b0..d71a93a 100644
--- a/.github/actions/start-services/action.yaml
+++ b/.github/actions/start-services/action.yaml
@@ -1,5 +1,5 @@
 name: Start Services
-description: Checkout the repo and start Docker Compose services
+description: Start Docker Compose services
 
 inputs:
   profiles:
@@ -14,9 +14,6 @@ inputs:
 runs:
   using: composite
   steps:
-    - name: Checkout
-      uses: actions/checkout@v4
-
     - name: Start services
       shell: bash
       run: |
diff --git a/.github/workflows/integration-test.yaml b/.github/workflows/integration-test.yaml
index 2dc8364..839182b 100644
--- a/.github/workflows/integration-test.yaml
+++ b/.github/workflows/integration-test.yaml
@@ -13,6 +13,9 @@ jobs:
     timeout-minutes: 15
 
     steps:
+      - name: Checkout
+        uses: actions/checkout@v4
+
       - name: Start services
         uses: ./.github/actions/start-services
         with:
@@ -28,11 +31,18 @@ jobs:
       - name: Dataset Processing
         run: bash scripts/wait_for_dataset.sh "${{ steps.dataset.outputs.id }}" 180
 
+      - name: Show logs on failure
+        if: failure()
+        run: docker compose logs --tail=50
+
   task-processing:
     runs-on: ubuntu-latest
     timeout-minutes: 15
 
     steps:
+      - name: Checkout
+        uses: actions/checkout@v4
+
       - name: Start services
         uses: ./.github/actions/start-services
         with:

From 281fdb464d2f53acd6ad0c0025461829a61d2584 Mon Sep 17 00:00:00 2001
From: PGijsbers <p.gijsbers@tue.nl>
Date: Thu, 8 Oct 2026 11:58:53 +0200
Subject: [PATCH 05/12] Remove timeout directive as it's not allowed in comp
 actions

---
 .github/actions/start-services/action.yaml | 1 -
 1 file changed, 1 deletion(-)

diff --git a/.github/actions/start-services/action.yaml b/.github/actions/start-services/action.yaml
index d71a93a..94665f2 100644
--- a/.github/actions/start-services/action.yaml
+++ b/.github/actions/start-services/action.yaml
@@ -23,7 +23,6 @@ runs:
         done
         SERVICES=$(echo "${{ inputs.services }}" | tr ',' ' ')
         docker compose $PROFILE_FLAGS up -d --build --wait $SERVICES
-      timeout-minutes: 10
 
     - name: Show service status
       if: always()

From cdb2292a266c57328684967bb1f1c03aace404ef Mon Sep 17 00:00:00 2001
From: PGijsbers <p.gijsbers@tue.nl>
Date: Thu, 8 Oct 2026 12:10:21 +0200
Subject: [PATCH 06/12] Set permissions to allow containers to use the mounted
 volumes

---
 .github/actions/start-services/action.yaml | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/.github/actions/start-services/action.yaml b/.github/actions/start-services/action.yaml
index 94665f2..1e8aad5 100644
--- a/.github/actions/start-services/action.yaml
+++ b/.github/actions/start-services/action.yaml
@@ -14,6 +14,11 @@ inputs:
 runs:
   using: composite
   steps:
+    - name: Set permission for mount paths
+      shell: bash
+      run: |
+        chmod -R 777 ./data/php
+        chmod -R 777 ./logs
     - name: Start services
       shell: bash
       run: |

From 3eb2a216a7467c8a7e80de82ddb7da3900b59421 Mon Sep 17 00:00:00 2001
From: PGijsbers <p.gijsbers@tue.nl>
Date: Thu, 8 Oct 2026 12:33:45 +0200
Subject: [PATCH 07/12] Print body on error

---
 scripts/wait_for_task.sh | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/scripts/wait_for_task.sh b/scripts/wait_for_task.sh
index 4f9d3e6..6fdc98b 100755
--- a/scripts/wait_for_task.sh
+++ b/scripts/wait_for_task.sh
@@ -13,12 +13,15 @@ while [ "$ELAPSED" -lt "$TIMEOUT" ]; do
   SPLITS_URL=$(echo "$RESPONSE" | sed -n 's/.*<oml:data_splits_url>\(.*\)<\/oml:data_splits_url>.*/\1/p')
 
   if [ -n "$SPLITS_URL" ]; then
-    STATUS_CODE=$(curl -s -o /dev/null -w "%{http_code}" "$SPLITS_URL")
+    SPLITS_RESPONSE=$(curl -s -w "\n%{http_code}" "$SPLITS_URL")
+    STATUS_CODE=$(echo "$SPLITS_RESPONSE" | tail -n1)
+    SPLITS_BODY=$(echo "$SPLITS_RESPONSE" | sed '$d')
     if [ "$STATUS_CODE" = "200" ]; then
       echo "Task $TASK_ID has split files available."
       exit 0
     fi
     echo "Splits URL exists but returned HTTP $STATUS_CODE after ${ELAPSED}s"
+    echo "$SPLITS_BODY"
   else
     echo "No splits URL yet after ${ELAPSED}s"
   fi

From 489cc1c2619ac75d97bf2a141783e2519c01a215 Mon Sep 17 00:00:00 2001
From: PGijsbers <p.gijsbers@tue.nl>
Date: Thu, 8 Oct 2026 13:44:47 +0200
Subject: [PATCH 08/12] Start working on run evaluation check

---
 .github/workflows/integration-test.yaml |  27 +++
 scripts/data/run/description.xml        | 305 ++++++++++++++++++++++++
 scripts/data/run/predictions.arff       | 170 +++++++++++++
 scripts/upload_run.sh                   |  20 ++
 scripts/wait_for_run.sh                 |  26 ++
 5 files changed, 548 insertions(+)
 create mode 100644 scripts/data/run/description.xml
 create mode 100644 scripts/data/run/predictions.arff
 create mode 100755 scripts/upload_run.sh
 create mode 100755 scripts/wait_for_run.sh

diff --git a/.github/workflows/integration-test.yaml b/.github/workflows/integration-test.yaml
index 839182b..a01771b 100644
--- a/.github/workflows/integration-test.yaml
+++ b/.github/workflows/integration-test.yaml
@@ -61,3 +61,30 @@ jobs:
       - name: Show logs on failure
         if: failure()
         run: docker compose logs --tail=50
+
+  run-processing:
+    runs-on: ubuntu-latest
+    timeout-minutes: 15
+
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v4
+
+      - name: Start services
+        uses: ./.github/actions/start-services
+        with:
+          profiles: "rest-api,evaluation-engine"
+
+      - name: Upload run
+        id: run
+        run: |
+          RUN_ID=$(bash scripts/upload_run.sh)
+          echo "Run ID: $RUN_ID"
+          echo "id=$RUN_ID" >> "$GITHUB_OUTPUT"
+
+      - name: Run Processing
+        run: bash scripts/wait_for_run.sh "${{ steps.run.outputs.id }}" 180
+
+      - name: Show logs on failure
+        if: failure()
+        run: docker compose logs --tail=50
diff --git a/scripts/data/run/description.xml b/scripts/data/run/description.xml
new file mode 100644
index 0000000..032178e
--- /dev/null
+++ b/scripts/data/run/description.xml
@@ -0,0 +1,305 @@
+<oml:run xmlns:oml="http://openml.org/openml">
+	<oml:task_id>1196</oml:task_id>
+	<oml:flow_id>1</oml:flow_id>
+	<oml:setup_string>Python_3.10.16. Sklearn_1.6.1. NumPy_2.2.2. SciPy_1.15.1.</oml:setup_string>
+  <oml:parameter_setting>
+        <oml:name>-do-not-check-capabilities</oml:name>
+        <oml:value></oml:value>
+  </oml:parameter_setting>
+  <oml:parameter_setting>
+        <oml:name>batch-size</oml:name>
+        <oml:value>100</oml:value>
+  </oml:parameter_setting>
+  <oml:parameter_setting>
+        <oml:name>num-decimal-places</oml:name>
+        <oml:value>2</oml:value>
+  </oml:parameter_setting>
+  <oml:parameter_setting>
+        <oml:name>output-debug-info</oml:name>
+        <oml:value></oml:value>
+  </oml:parameter_setting>
+	<oml:tag>openml-python</oml:tag>
+	<oml:tag>Sklearn_1.6.1.</oml:tag>
+	<oml:output_data>
+		<oml:evaluation repeat="0" fold="0">
+			<oml:name>usercpu_time_millis_training</oml:name>
+			<oml:value>11.353665999999762</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="1">
+			<oml:name>usercpu_time_millis_training</oml:name>
+			<oml:value>1.2970839999999484</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="2">
+			<oml:name>usercpu_time_millis_training</oml:name>
+			<oml:value>1.1723329999995258</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="3">
+			<oml:name>usercpu_time_millis_training</oml:name>
+			<oml:value>1.1706670000002362</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="4">
+			<oml:name>usercpu_time_millis_training</oml:name>
+			<oml:value>1.2168329999999727</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="5">
+			<oml:name>usercpu_time_millis_training</oml:name>
+			<oml:value>1.226291000000046</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="6">
+			<oml:name>usercpu_time_millis_training</oml:name>
+			<oml:value>1.2974170000008556</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="7">
+			<oml:name>usercpu_time_millis_training</oml:name>
+			<oml:value>1.2798339999999797</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="8">
+			<oml:name>usercpu_time_millis_training</oml:name>
+			<oml:value>1.1370409999997833</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="9">
+			<oml:name>usercpu_time_millis_training</oml:name>
+			<oml:value>1.3173340000003364</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="0">
+			<oml:name>wall_clock_time_millis_training</oml:name>
+			<oml:value>11.349678039550781</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="1">
+			<oml:name>wall_clock_time_millis_training</oml:name>
+			<oml:value>1.294851303100586</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="2">
+			<oml:name>wall_clock_time_millis_training</oml:name>
+			<oml:value>1.1701583862304688</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="3">
+			<oml:name>wall_clock_time_millis_training</oml:name>
+			<oml:value>1.1692047119140625</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="4">
+			<oml:name>wall_clock_time_millis_training</oml:name>
+			<oml:value>1.2137889862060547</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="5">
+			<oml:name>wall_clock_time_millis_training</oml:name>
+			<oml:value>1.2259483337402344</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="6">
+			<oml:name>wall_clock_time_millis_training</oml:name>
+			<oml:value>1.294851303100586</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="7">
+			<oml:name>wall_clock_time_millis_training</oml:name>
+			<oml:value>1.2731552124023438</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="8">
+			<oml:name>wall_clock_time_millis_training</oml:name>
+			<oml:value>1.1365413665771484</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="9">
+			<oml:name>wall_clock_time_millis_training</oml:name>
+			<oml:value>1.318216323852539</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="0">
+			<oml:name>usercpu_time_millis_testing</oml:name>
+			<oml:value>0.8559159999999011</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="1">
+			<oml:name>usercpu_time_millis_testing</oml:name>
+			<oml:value>0.4579580000001471</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="2">
+			<oml:name>usercpu_time_millis_testing</oml:name>
+			<oml:value>0.45920799999965567</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="3">
+			<oml:name>usercpu_time_millis_testing</oml:name>
+			<oml:value>0.449209000000117</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="4">
+			<oml:name>usercpu_time_millis_testing</oml:name>
+			<oml:value>0.452333000000138</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="5">
+			<oml:name>usercpu_time_millis_testing</oml:name>
+			<oml:value>0.4404999999998438</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="6">
+			<oml:name>usercpu_time_millis_testing</oml:name>
+			<oml:value>0.46437499999996135</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="7">
+			<oml:name>usercpu_time_millis_testing</oml:name>
+			<oml:value>0.43887499999950563</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="8">
+			<oml:name>usercpu_time_millis_testing</oml:name>
+			<oml:value>0.4358330000000521</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="9">
+			<oml:name>usercpu_time_millis_testing</oml:name>
+			<oml:value>0.4521660000005312</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="0">
+			<oml:name>usercpu_time_millis</oml:name>
+			<oml:value>12.209581999999664</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="1">
+			<oml:name>usercpu_time_millis</oml:name>
+			<oml:value>1.7550420000000955</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="2">
+			<oml:name>usercpu_time_millis</oml:name>
+			<oml:value>1.6315409999991815</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="3">
+			<oml:name>usercpu_time_millis</oml:name>
+			<oml:value>1.6198760000003531</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="4">
+			<oml:name>usercpu_time_millis</oml:name>
+			<oml:value>1.6691660000001107</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="5">
+			<oml:name>usercpu_time_millis</oml:name>
+			<oml:value>1.6667909999998898</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="6">
+			<oml:name>usercpu_time_millis</oml:name>
+			<oml:value>1.761792000000817</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="7">
+			<oml:name>usercpu_time_millis</oml:name>
+			<oml:value>1.7187089999994853</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="8">
+			<oml:name>usercpu_time_millis</oml:name>
+			<oml:value>1.5728739999998353</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="9">
+			<oml:name>usercpu_time_millis</oml:name>
+			<oml:value>1.7695000000008676</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="0">
+			<oml:name>wall_clock_time_millis_testing</oml:name>
+			<oml:value>0.8571147918701172</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="1">
+			<oml:name>wall_clock_time_millis_testing</oml:name>
+			<oml:value>0.45943260192871094</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="2">
+			<oml:name>wall_clock_time_millis_testing</oml:name>
+			<oml:value>0.4608631134033203</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="3">
+			<oml:name>wall_clock_time_millis_testing</oml:name>
+			<oml:value>0.4506111145019531</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="4">
+			<oml:name>wall_clock_time_millis_testing</oml:name>
+			<oml:value>0.45418739318847656</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="5">
+			<oml:name>wall_clock_time_millis_testing</oml:name>
+			<oml:value>0.44226646423339844</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="6">
+			<oml:name>wall_clock_time_millis_testing</oml:name>
+			<oml:value>0.46944618225097656</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="7">
+			<oml:name>wall_clock_time_millis_testing</oml:name>
+			<oml:value>0.44035911560058594</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="8">
+			<oml:name>wall_clock_time_millis_testing</oml:name>
+			<oml:value>0.4374980926513672</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="9">
+			<oml:name>wall_clock_time_millis_testing</oml:name>
+			<oml:value>0.45371055603027344</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="0">
+			<oml:name>wall_clock_time_millis</oml:name>
+			<oml:value>12.206792831420898</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="1">
+			<oml:name>wall_clock_time_millis</oml:name>
+			<oml:value>1.7542839050292969</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="2">
+			<oml:name>wall_clock_time_millis</oml:name>
+			<oml:value>1.631021499633789</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="3">
+			<oml:name>wall_clock_time_millis</oml:name>
+			<oml:value>1.6198158264160156</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="4">
+			<oml:name>wall_clock_time_millis</oml:name>
+			<oml:value>1.6679763793945312</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="5">
+			<oml:name>wall_clock_time_millis</oml:name>
+			<oml:value>1.6682147979736328</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="6">
+			<oml:name>wall_clock_time_millis</oml:name>
+			<oml:value>1.7642974853515625</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="7">
+			<oml:name>wall_clock_time_millis</oml:name>
+			<oml:value>1.7135143280029297</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="8">
+			<oml:name>wall_clock_time_millis</oml:name>
+			<oml:value>1.5740394592285156</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="9">
+			<oml:name>wall_clock_time_millis</oml:name>
+			<oml:value>1.7719268798828125</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="0">
+			<oml:name>predictive_accuracy</oml:name>
+			<oml:value>0.9333333333333333</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="1">
+			<oml:name>predictive_accuracy</oml:name>
+			<oml:value>1.0</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="2">
+			<oml:name>predictive_accuracy</oml:name>
+			<oml:value>1.0</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="3">
+			<oml:name>predictive_accuracy</oml:name>
+			<oml:value>0.9333333333333333</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="4">
+			<oml:name>predictive_accuracy</oml:name>
+			<oml:value>1.0</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="5">
+			<oml:name>predictive_accuracy</oml:name>
+			<oml:value>0.9333333333333333</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="6">
+			<oml:name>predictive_accuracy</oml:name>
+			<oml:value>1.0</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="7">
+			<oml:name>predictive_accuracy</oml:name>
+			<oml:value>0.8666666666666667</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="8">
+			<oml:name>predictive_accuracy</oml:name>
+			<oml:value>0.8666666666666667</oml:value>
+		</oml:evaluation>
+		<oml:evaluation repeat="0" fold="9">
+			<oml:name>predictive_accuracy</oml:name>
+			<oml:value>1.0</oml:value>
+		</oml:evaluation>
+	</oml:output_data>
+</oml:run>
diff --git a/scripts/data/run/predictions.arff b/scripts/data/run/predictions.arff
new file mode 100644
index 0000000..31a0eb3
--- /dev/null
+++ b/scripts/data/run/predictions.arff
@@ -0,0 +1,170 @@
+% Python_3.10.16.
+% Sklearn_1.6.1.
+% NumPy_2.2.2.
+% SciPy_1.15.1.
+% Thu Oct  8 10:27:27 2026
+% Created by run_flow_on_task
+@RELATION openml_task_1196_predictions
+
+@ATTRIBUTE repeat NUMERIC
+@ATTRIBUTE fold NUMERIC
+@ATTRIBUTE sample NUMERIC
+@ATTRIBUTE row_id NUMERIC
+@ATTRIBUTE prediction {Iris-setosa, Iris-versicolor, Iris-virginica}
+@ATTRIBUTE correct {Iris-setosa, Iris-versicolor, Iris-virginica}
+@ATTRIBUTE confidence.Iris-setosa NUMERIC
+@ATTRIBUTE confidence.Iris-versicolor NUMERIC
+@ATTRIBUTE confidence.Iris-virginica NUMERIC
+
+@DATA
+0,0,0,43,Iris-setosa,Iris-setosa,1.0,0.0,0.0
+0,0,0,14,Iris-setosa,Iris-setosa,1.0,0.0,0.0
+0,0,0,37,Iris-setosa,Iris-setosa,1.0,0.0,0.0
+0,0,0,23,Iris-setosa,Iris-setosa,1.0,0.0,0.0
+0,0,0,10,Iris-setosa,Iris-setosa,1.0,0.0,0.0
+0,0,0,99,Iris-versicolor,Iris-versicolor,0.0,1.0,0.0
+0,0,0,87,Iris-versicolor,Iris-versicolor,0.0,1.0,0.0
+0,0,0,97,Iris-versicolor,Iris-versicolor,0.0,1.0,0.0
+0,0,0,62,Iris-versicolor,Iris-versicolor,0.0,1.0,0.0
+0,0,0,92,Iris-versicolor,Iris-versicolor,0.0,1.0,0.0
+0,0,0,119,Iris-versicolor,Iris-virginica,0.0,1.0,0.0
+0,0,0,111,Iris-virginica,Iris-virginica,0.0,0.0,1.0
+0,0,0,144,Iris-virginica,Iris-virginica,0.0,0.0,1.0
+0,0,0,116,Iris-virginica,Iris-virginica,0.0,0.0,1.0
+0,0,0,125,Iris-virginica,Iris-virginica,0.0,0.0,1.0
+0,1,0,49,Iris-setosa,Iris-setosa,1.0,0.0,0.0
+0,1,0,15,Iris-setosa,Iris-setosa,1.0,0.0,0.0
+0,1,0,47,Iris-setosa,Iris-setosa,1.0,0.0,0.0
+0,1,0,0,Iris-setosa,Iris-setosa,1.0,0.0,0.0
+0,1,0,44,Iris-setosa,Iris-setosa,1.0,0.0,0.0
+0,1,0,76,Iris-versicolor,Iris-versicolor,0.0,1.0,0.0
+0,1,0,89,Iris-versicolor,Iris-versicolor,0.0,1.0,0.0
+0,1,0,58,Iris-versicolor,Iris-versicolor,0.0,1.0,0.0
+0,1,0,54,Iris-versicolor,Iris-versicolor,0.0,1.0,0.0
+0,1,0,53,Iris-versicolor,Iris-versicolor,0.0,1.0,0.0
+0,1,0,149,Iris-virginica,Iris-virginica,0.0,0.0,1.0
+0,1,0,114,Iris-virginica,Iris-virginica,0.0,0.0,1.0
+0,1,0,112,Iris-virginica,Iris-virginica,0.0,0.0,1.0
+0,1,0,115,Iris-virginica,Iris-virginica,0.0,0.0,1.0
+0,1,0,139,Iris-virginica,Iris-virginica,0.0,0.0,1.0
+0,2,0,42,Iris-setosa,Iris-setosa,1.0,0.0,0.0
+0,2,0,41,Iris-setosa,Iris-setosa,1.0,0.0,0.0
+0,2,0,31,Iris-setosa,Iris-setosa,1.0,0.0,0.0
+0,2,0,48,Iris-setosa,Iris-setosa,1.0,0.0,0.0
+0,2,0,7,Iris-setosa,Iris-setosa,1.0,0.0,0.0
+0,2,0,94,Iris-versicolor,Iris-versicolor,0.0,1.0,0.0
+0,2,0,93,Iris-versicolor,Iris-versicolor,0.0,1.0,0.0
+0,2,0,95,Iris-versicolor,Iris-versicolor,0.0,1.0,0.0
+0,2,0,51,Iris-versicolor,Iris-versicolor,0.0,1.0,0.0
+0,2,0,72,Iris-versicolor,Iris-versicolor,0.0,1.0,0.0
+0,2,0,135,Iris-virginica,Iris-virginica,0.0,0.0,1.0
+0,2,0,124,Iris-virginica,Iris-virginica,0.0,0.0,1.0
+0,2,0,141,Iris-virginica,Iris-virginica,0.0,0.0,1.0
+0,2,0,105,Iris-virginica,Iris-virginica,0.0,0.0,1.0
+0,2,0,143,Iris-virginica,Iris-virginica,0.0,0.0,1.0
+0,3,0,34,Iris-setosa,Iris-setosa,1.0,0.0,0.0
+0,3,0,22,Iris-setosa,Iris-setosa,1.0,0.0,0.0
+0,3,0,45,Iris-setosa,Iris-setosa,1.0,0.0,0.0
+0,3,0,3,Iris-setosa,Iris-setosa,1.0,0.0,0.0
+0,3,0,16,Iris-setosa,Iris-setosa,1.0,0.0,0.0
+0,3,0,79,Iris-versicolor,Iris-versicolor,0.0,1.0,0.0
+0,3,0,55,Iris-versicolor,Iris-versicolor,0.0,1.0,0.0
+0,3,0,98,Iris-versicolor,Iris-versicolor,0.0,1.0,0.0
+0,3,0,80,Iris-versicolor,Iris-versicolor,0.0,1.0,0.0
+0,3,0,77,Iris-virginica,Iris-versicolor,0.0,0.0,1.0
+0,3,0,120,Iris-virginica,Iris-virginica,0.0,0.0,1.0
+0,3,0,122,Iris-virginica,Iris-virginica,0.0,0.0,1.0
+0,3,0,145,Iris-virginica,Iris-virginica,0.0,0.0,1.0
+0,3,0,100,Iris-virginica,Iris-virginica,0.0,0.0,1.0
+0,3,0,107,Iris-virginica,Iris-virginica,0.0,0.0,1.0
+0,4,0,4,Iris-setosa,Iris-setosa,1.0,0.0,0.0
+0,4,0,27,Iris-setosa,Iris-setosa,1.0,0.0,0.0
+0,4,0,38,Iris-setosa,Iris-setosa,1.0,0.0,0.0
+0,4,0,19,Iris-setosa,Iris-setosa,1.0,0.0,0.0
+0,4,0,33,Iris-setosa,Iris-setosa,1.0,0.0,0.0
+0,4,0,96,Iris-versicolor,Iris-versicolor,0.0,1.0,0.0
+0,4,0,50,Iris-versicolor,Iris-versicolor,0.0,1.0,0.0
+0,4,0,61,Iris-versicolor,Iris-versicolor,0.0,1.0,0.0
+0,4,0,65,Iris-versicolor,Iris-versicolor,0.0,1.0,0.0
+0,4,0,73,Iris-versicolor,Iris-versicolor,0.0,1.0,0.0
+0,4,0,109,Iris-virginica,Iris-virginica,0.0,0.0,1.0
+0,4,0,126,Iris-virginica,Iris-virginica,0.0,0.0,1.0
+0,4,0,110,Iris-virginica,Iris-virginica,0.0,0.0,1.0
+0,4,0,148,Iris-virginica,Iris-virginica,0.0,0.0,1.0
+0,4,0,103,Iris-virginica,Iris-virginica,0.0,0.0,1.0
+0,5,0,46,Iris-setosa,Iris-setosa,1.0,0.0,0.0
+0,5,0,6,Iris-setosa,Iris-setosa,1.0,0.0,0.0
+0,5,0,18,Iris-setosa,Iris-setosa,1.0,0.0,0.0
+0,5,0,36,Iris-setosa,Iris-setosa,1.0,0.0,0.0
+0,5,0,24,Iris-setosa,Iris-setosa,1.0,0.0,0.0
+0,5,0,83,Iris-virginica,Iris-versicolor,0.0,0.0,1.0
+0,5,0,90,Iris-versicolor,Iris-versicolor,0.0,1.0,0.0
+0,5,0,68,Iris-versicolor,Iris-versicolor,0.0,1.0,0.0
+0,5,0,52,Iris-versicolor,Iris-versicolor,0.0,1.0,0.0
+0,5,0,75,Iris-versicolor,Iris-versicolor,0.0,1.0,0.0
+0,5,0,134,Iris-virginica,Iris-virginica,0.0,0.0,1.0
+0,5,0,104,Iris-virginica,Iris-virginica,0.0,0.0,1.0
+0,5,0,117,Iris-virginica,Iris-virginica,0.0,0.0,1.0
+0,5,0,127,Iris-virginica,Iris-virginica,0.0,0.0,1.0
+0,5,0,121,Iris-virginica,Iris-virginica,0.0,0.0,1.0
+0,6,0,26,Iris-setosa,Iris-setosa,1.0,0.0,0.0
+0,6,0,13,Iris-setosa,Iris-setosa,1.0,0.0,0.0
+0,6,0,32,Iris-setosa,Iris-setosa,1.0,0.0,0.0
+0,6,0,20,Iris-setosa,Iris-setosa,1.0,0.0,0.0
+0,6,0,17,Iris-setosa,Iris-setosa,1.0,0.0,0.0
+0,6,0,63,Iris-versicolor,Iris-versicolor,0.0,1.0,0.0
+0,6,0,88,Iris-versicolor,Iris-versicolor,0.0,1.0,0.0
+0,6,0,64,Iris-versicolor,Iris-versicolor,0.0,1.0,0.0
+0,6,0,91,Iris-versicolor,Iris-versicolor,0.0,1.0,0.0
+0,6,0,81,Iris-versicolor,Iris-versicolor,0.0,1.0,0.0
+0,6,0,137,Iris-virginica,Iris-virginica,0.0,0.0,1.0
+0,6,0,101,Iris-virginica,Iris-virginica,0.0,0.0,1.0
+0,6,0,130,Iris-virginica,Iris-virginica,0.0,0.0,1.0
+0,6,0,118,Iris-virginica,Iris-virginica,0.0,0.0,1.0
+0,6,0,108,Iris-virginica,Iris-virginica,0.0,0.0,1.0
+0,7,0,29,Iris-setosa,Iris-setosa,1.0,0.0,0.0
+0,7,0,40,Iris-setosa,Iris-setosa,1.0,0.0,0.0
+0,7,0,12,Iris-setosa,Iris-setosa,1.0,0.0,0.0
+0,7,0,35,Iris-setosa,Iris-setosa,1.0,0.0,0.0
+0,7,0,8,Iris-setosa,Iris-setosa,1.0,0.0,0.0
+0,7,0,86,Iris-versicolor,Iris-versicolor,0.0,1.0,0.0
+0,7,0,82,Iris-versicolor,Iris-versicolor,0.0,1.0,0.0
+0,7,0,84,Iris-versicolor,Iris-versicolor,0.0,1.0,0.0
+0,7,0,74,Iris-versicolor,Iris-versicolor,0.0,1.0,0.0
+0,7,0,57,Iris-versicolor,Iris-versicolor,0.0,1.0,0.0
+0,7,0,140,Iris-virginica,Iris-virginica,0.0,0.0,1.0
+0,7,0,133,Iris-versicolor,Iris-virginica,0.0,1.0,0.0
+0,7,0,106,Iris-versicolor,Iris-virginica,0.0,1.0,0.0
+0,7,0,128,Iris-virginica,Iris-virginica,0.0,0.0,1.0
+0,7,0,102,Iris-virginica,Iris-virginica,0.0,0.0,1.0
+0,8,0,25,Iris-setosa,Iris-setosa,1.0,0.0,0.0
+0,8,0,1,Iris-setosa,Iris-setosa,1.0,0.0,0.0
+0,8,0,21,Iris-setosa,Iris-setosa,1.0,0.0,0.0
+0,8,0,30,Iris-setosa,Iris-setosa,1.0,0.0,0.0
+0,8,0,28,Iris-setosa,Iris-setosa,1.0,0.0,0.0
+0,8,0,67,Iris-versicolor,Iris-versicolor,0.0,1.0,0.0
+0,8,0,78,Iris-versicolor,Iris-versicolor,0.0,1.0,0.0
+0,8,0,66,Iris-versicolor,Iris-versicolor,0.0,1.0,0.0
+0,8,0,70,Iris-virginica,Iris-versicolor,0.0,0.0,1.0
+0,8,0,59,Iris-versicolor,Iris-versicolor,0.0,1.0,0.0
+0,8,0,136,Iris-virginica,Iris-virginica,0.0,0.0,1.0
+0,8,0,138,Iris-virginica,Iris-virginica,0.0,0.0,1.0
+0,8,0,129,Iris-versicolor,Iris-virginica,0.0,1.0,0.0
+0,8,0,113,Iris-virginica,Iris-virginica,0.0,0.0,1.0
+0,8,0,132,Iris-virginica,Iris-virginica,0.0,0.0,1.0
+0,9,0,5,Iris-setosa,Iris-setosa,1.0,0.0,0.0
+0,9,0,11,Iris-setosa,Iris-setosa,1.0,0.0,0.0
+0,9,0,9,Iris-setosa,Iris-setosa,1.0,0.0,0.0
+0,9,0,39,Iris-setosa,Iris-setosa,1.0,0.0,0.0
+0,9,0,2,Iris-setosa,Iris-setosa,1.0,0.0,0.0
+0,9,0,85,Iris-versicolor,Iris-versicolor,0.0,1.0,0.0
+0,9,0,56,Iris-versicolor,Iris-versicolor,0.0,1.0,0.0
+0,9,0,69,Iris-versicolor,Iris-versicolor,0.0,1.0,0.0
+0,9,0,71,Iris-versicolor,Iris-versicolor,0.0,1.0,0.0
+0,9,0,60,Iris-versicolor,Iris-versicolor,0.0,1.0,0.0
+0,9,0,142,Iris-virginica,Iris-virginica,0.0,0.0,1.0
+0,9,0,123,Iris-virginica,Iris-virginica,0.0,0.0,1.0
+0,9,0,131,Iris-virginica,Iris-virginica,0.0,0.0,1.0
+0,9,0,147,Iris-virginica,Iris-virginica,0.0,0.0,1.0
+0,9,0,146,Iris-virginica,Iris-virginica,0.0,0.0,1.0
+
diff --git a/scripts/upload_run.sh b/scripts/upload_run.sh
new file mode 100755
index 0000000..5b7a561
--- /dev/null
+++ b/scripts/upload_run.sh
@@ -0,0 +1,20 @@
+#!/bin/bash
+# Uploads a run to the REST API and prints the run ID to stdout.
+# The predictions and description files are fixed in scripts/data/run/.
+set -euo pipefail
+
+SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
+
+RESPONSE=$(curl -s -X POST "http://localhost:8000/api/v1/xml/run?api_key=normaluser" \
+  -F "description=@${SCRIPT_DIR}/data/run/description.xml;type=text/xml" \
+  -F "predictions=@${SCRIPT_DIR}/data/run/predictions.arff")
+
+RUN_ID=$(echo "$RESPONSE" | sed -n 's/.*<oml:id>\([0-9]*\)<\/oml:id>.*/\1/p')
+
+if [ -z "$RUN_ID" ]; then
+  echo "Server response:" >&2
+  echo "$RESPONSE" >&2
+  exit 1
+fi
+
+echo "$RUN_ID"
diff --git a/scripts/wait_for_run.sh b/scripts/wait_for_run.sh
new file mode 100755
index 0000000..29c273e
--- /dev/null
+++ b/scripts/wait_for_run.sh
@@ -0,0 +1,26 @@
+#!/bin/bash
+# Waits until a run has evaluations computed by the evaluation engine.
+# Usage: wait_for_run.sh RUN_ID [TIMEOUT]
+set -euo pipefail
+
+RUN_ID="${1:?Usage: wait_for_run.sh RUN_ID [TIMEOUT]}"
+TIMEOUT="${2:-120}"
+INTERVAL=5
+ELAPSED=0
+
+while [ "$ELAPSED" -lt "$TIMEOUT" ]; do
+  RESPONSE=$(curl -s "http://localhost:8000/api/v1/xml/run/${RUN_ID}")
+
+  if echo "$RESPONSE" | grep -q "area_under_roc_curve"; then
+    echo "Run $RUN_ID has evaluations."
+    exit 0
+  fi
+
+  echo "No evaluations yet after ${ELAPSED}s"
+
+  sleep "$INTERVAL"
+  ELAPSED=$((ELAPSED + INTERVAL))
+done
+
+echo "ERROR: Run $RUN_ID did not get evaluations within ${TIMEOUT} seconds."
+exit 1

From 047c24c010821099c76539121613d5ac5f5459c1 Mon Sep 17 00:00:00 2001
From: PGijsbers <p.gijsbers@tue.nl>
Date: Tue, 13 Oct 2026 12:16:41 +0200
Subject: [PATCH 09/12] Use correct identifier to extract run ID

---
 scripts/upload_run.sh | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/scripts/upload_run.sh b/scripts/upload_run.sh
index 5b7a561..27d2d43 100755
--- a/scripts/upload_run.sh
+++ b/scripts/upload_run.sh
@@ -9,7 +9,7 @@ RESPONSE=$(curl -s -X POST "http://localhost:8000/api/v1/xml/run?api_key=normalu
   -F "description=@${SCRIPT_DIR}/data/run/description.xml;type=text/xml" \
   -F "predictions=@${SCRIPT_DIR}/data/run/predictions.arff")
 
-RUN_ID=$(echo "$RESPONSE" | sed -n 's/.*<oml:id>\([0-9]*\)<\/oml:id>.*/\1/p')
+RUN_ID=$(echo "$RESPONSE" | sed -n 's/.*<oml:run_id>\([0-9]*\)<\/oml:run_id>.*/\1/p')
 
 if [ -z "$RUN_ID" ]; then
   echo "Server response:" >&2

From 3e923326080672bcb3cc98e139c9e8526143fed7 Mon Sep 17 00:00:00 2001
From: PGijsbers <p.gijsbers@tue.nl>
Date: Tue, 13 Oct 2026 16:02:59 +0200
Subject: [PATCH 10/12] remove unused file

---
 scripts/evaluation_engine_dataset.sh | 10 ----------
 1 file changed, 10 deletions(-)
 delete mode 100644 scripts/evaluation_engine_dataset.sh

diff --git a/scripts/evaluation_engine_dataset.sh b/scripts/evaluation_engine_dataset.sh
deleted file mode 100644
index dd4f2ee..0000000
--- a/scripts/evaluation_engine_dataset.sh
+++ /dev/null
@@ -1,10 +0,0 @@
-#!/bin/bash
-# Tests the dataset upload and processing of the REST API and Evaluation Engine
-set -euo pipefail
-
-SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
-
-DATASET_ID=$("${SCRIPT_DIR}/upload_dataset.sh")
-echo "Uploaded dataset with ID: $DATASET_ID"
-
-"${SCRIPT_DIR}/wait_for_dataset.sh" "$DATASET_ID"

From c88cd03fb77ce62c0e9b9de8b516cc2f6dc40f26 Mon Sep 17 00:00:00 2001
From: PGijsbers <p.gijsbers@tue.nl>
Date: Tue, 13 Oct 2026 16:05:01 +0200
Subject: [PATCH 11/12] show errors for curl

---
 scripts/create_task.sh      | 2 +-
 scripts/upload_dataset.sh   | 2 +-
 scripts/upload_run.sh       | 2 +-
 scripts/wait_for_dataset.sh | 2 +-
 scripts/wait_for_run.sh     | 2 +-
 scripts/wait_for_task.sh    | 4 ++--
 6 files changed, 7 insertions(+), 7 deletions(-)

diff --git a/scripts/create_task.sh b/scripts/create_task.sh
index d09bc99..d7bc9ac 100755
--- a/scripts/create_task.sh
+++ b/scripts/create_task.sh
@@ -19,7 +19,7 @@ cat > "$TMPFILE" << EOF
 </oml:task_inputs>
 EOF
 
-RESPONSE=$(curl -s -X POST "http://localhost:8000/api/v1/xml/task?api_key=normaluser" \
+RESPONSE=$(curl -sS -X POST "http://localhost:8000/api/v1/xml/task?api_key=normaluser" \
   -F "description=@${TMPFILE};type=text/xml")
 
 TASK_ID=$(echo "$RESPONSE" | sed -n 's/.*<oml:id>\([0-9]*\)<\/oml:id>.*/\1/p')
diff --git a/scripts/upload_dataset.sh b/scripts/upload_dataset.sh
index e031ad6..669ceb1 100755
--- a/scripts/upload_dataset.sh
+++ b/scripts/upload_dataset.sh
@@ -23,7 +23,7 @@ TMPFILE=$(mktemp /tmp/dataset_desc.XXXXXX.xml)
 trap 'rm -f "$TMPFILE"' EXIT
 echo "$XML_DESCRIPTION" > "$TMPFILE"
 
-RESPONSE=$(curl -s -X POST "http://localhost:8000/api/v1/xml/data?api_key=normaluser" \
+RESPONSE=$(curl -sS -X POST "http://localhost:8000/api/v1/xml/data?api_key=normaluser" \
   -F "description=@${TMPFILE};type=text/xml" \
   -F "dataset=@${SCRIPT_DIR}/data/test.arff")
 
diff --git a/scripts/upload_run.sh b/scripts/upload_run.sh
index 27d2d43..912b072 100755
--- a/scripts/upload_run.sh
+++ b/scripts/upload_run.sh
@@ -5,7 +5,7 @@ set -euo pipefail
 
 SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
 
-RESPONSE=$(curl -s -X POST "http://localhost:8000/api/v1/xml/run?api_key=normaluser" \
+RESPONSE=$(curl -sS -X POST "http://localhost:8000/api/v1/xml/run?api_key=normaluser" \
   -F "description=@${SCRIPT_DIR}/data/run/description.xml;type=text/xml" \
   -F "predictions=@${SCRIPT_DIR}/data/run/predictions.arff")
 
diff --git a/scripts/wait_for_dataset.sh b/scripts/wait_for_dataset.sh
index 6bdd0a9..f7c1c52 100755
--- a/scripts/wait_for_dataset.sh
+++ b/scripts/wait_for_dataset.sh
@@ -9,7 +9,7 @@ INTERVAL=5
 ELAPSED=0
 
 while [ "$ELAPSED" -lt "$TIMEOUT" ]; do
-  STATUS_RESPONSE=$(curl -s "http://localhost:8000/api/v1/xml/data/${DATASET_ID}")
+  STATUS_RESPONSE=$(curl -sS "http://localhost:8000/api/v1/xml/data/${DATASET_ID}")
   STATUS=$(echo "$STATUS_RESPONSE" | sed -n 's/.*<oml:status>\([a-zA-Z_]*\)<\/oml:status>.*/\1/p')
 
   echo "Status after ${ELAPSED}s: ${STATUS:-unknown}"
diff --git a/scripts/wait_for_run.sh b/scripts/wait_for_run.sh
index 29c273e..9cecbd4 100755
--- a/scripts/wait_for_run.sh
+++ b/scripts/wait_for_run.sh
@@ -9,7 +9,7 @@ INTERVAL=5
 ELAPSED=0
 
 while [ "$ELAPSED" -lt "$TIMEOUT" ]; do
-  RESPONSE=$(curl -s "http://localhost:8000/api/v1/xml/run/${RUN_ID}")
+  RESPONSE=$(curl -sS "http://localhost:8000/api/v1/xml/run/${RUN_ID}")
 
   if echo "$RESPONSE" | grep -q "area_under_roc_curve"; then
     echo "Run $RUN_ID has evaluations."
diff --git a/scripts/wait_for_task.sh b/scripts/wait_for_task.sh
index 6fdc98b..b77aece 100755
--- a/scripts/wait_for_task.sh
+++ b/scripts/wait_for_task.sh
@@ -9,11 +9,11 @@ INTERVAL=5
 ELAPSED=0
 
 while [ "$ELAPSED" -lt "$TIMEOUT" ]; do
-  RESPONSE=$(curl -s "http://localhost:8000/api/v1/xml/task/${TASK_ID}")
+  RESPONSE=$(curl -sS "http://localhost:8000/api/v1/xml/task/${TASK_ID}")
   SPLITS_URL=$(echo "$RESPONSE" | sed -n 's/.*<oml:data_splits_url>\(.*\)<\/oml:data_splits_url>.*/\1/p')
 
   if [ -n "$SPLITS_URL" ]; then
-    SPLITS_RESPONSE=$(curl -s -w "\n%{http_code}" "$SPLITS_URL")
+    SPLITS_RESPONSE=$(curl -sS -w "\n%{http_code}" "$SPLITS_URL")
     STATUS_CODE=$(echo "$SPLITS_RESPONSE" | tail -n1)
     SPLITS_BODY=$(echo "$SPLITS_RESPONSE" | sed '$d')
     if [ "$STATUS_CODE" = "200" ]; then

From 3a55ba166c431756caa63d4754b37b7f1d3c06b0 Mon Sep 17 00:00:00 2001
From: PGijsbers <p.gijsbers@tue.nl>
Date: Tue, 13 Oct 2026 16:13:02 +0200
Subject: [PATCH 12/12] Mention the new test scripts in the README

---
 README.md | 14 ++++++++++++++
 1 file changed, 14 insertions(+)

diff --git a/README.md b/README.md
index 9d76501..f18ab25 100644
--- a/README.md
+++ b/README.md
@@ -114,6 +114,20 @@ be slightly altered to use them in the test setup: change https to http.
 
 ## Development
 
+### Services
+When working on the `services` repository itself, the `scripts` directory provides a small number of integration tests to catch the most glaring issues. They are currently limited to REST API and evaluation engine functionality and also run automatically in CI.
+
+For example, after starting services, to test dataset upload run:
+```bash
+./scripts/upload_dataset.sh
+```
+If successful, this will also print the dataset id assigned to the new dataset.
+To test if dataset processing works, we can then run:
+```bash
+./scripts/wait_for_dataset.sh DATASET_ID
+```
+where `DATASET_ID` is replaced by the identifier printed in the previous step.
+
 ### PHP, Parquet and Croissant converter
 If you want to do local development on containers that are part of the docker-compose, you want those containers to change based on your code. You should have the relevant code somewhere on your system. You will need to tell the docker-compose where to find it. You can do so by setting environment variables.