Major UX improvements for credit-scorer: simplified workflows, unified visualizations, and user-friendly defaults

credit-scorer/README.md

@@ -115,20 +115,37 @@ zenml alerter register slack_alerter \
 zenml stack update <STACK_NAME> -al slack_alerter
 ```
 
+5. Set up Modal secrets for deployment (optional, only needed with `--enable-slack` flag):
+
+```bash
+# Create Modal secret with Slack credentials for incident reporting
+modal secret create credit-scoring-secrets \
+    SLACK_BOT_TOKEN=<your_slack_bot_token> \
+    SLACK_CHANNEL_ID=<your_slack_channel_id>
+```
+
+> **Note:** The deployment pipeline uses Modal for cloud deployment. By default, Slack notifications are disabled for easier testing. The `credit-scoring-secrets` Modal secret stores the necessary Slack credentials for automated notifications when the deployed model API detects high or critical severity incidents.
+
+> **Enabling full compliance features:** For complete EU AI Act compliance incident reporting (Article 18), use the `--enable-slack` flag (e.g., `python run.py --deploy --enable-slack`). This requires the Modal secret to be configured with your Slack credentials for automated incident notifications.
+
 ## 📊 Running Pipelines
 
 ### Basic Commands
 
 ```bash
+# Run complete workflow (recommended)
+python run.py --all                      # Feature → Training → Deployment (auto-approved, no Slack)
+
 # Run individual pipelines
 python run.py --feature    # Feature engineering (Articles 10, 12)
 python run.py --train      # Model training (Articles 9, 11, 15)
 python run.py --deploy     # Deployment (Articles 14, 17, 18)
 
 # Pipeline options
-python run.py --train --auto-approve     # Skip manual approval steps
-python run.py --feature --no-cache       # Disable ZenML caching
+python run.py --all --no-cache           # Complete workflow without caching
+python run.py --all --manual-approve     # Complete workflow with manual approval steps
 python run.py --deploy --config-dir ./my-configs  # Custom config directory
+python run.py --all --enable-slack       # Complete workflow with Slack notifications (requires Modal secrets)
 ```
 
 ### View Compliance Dashboard

credit-scorer/modal_app/modal_deployment.py

@@ -81,9 +81,33 @@ def create_modal_app(python_version: str = "3.12.9"):
 
     app_config = {
         "image": base_image,
-        "secrets": [modal.Secret.from_name(SECRET_NAME)],
     }
 
+    # Only add secrets if Slack notifications are explicitly enabled
+    enable_slack_raw = os.getenv("ENABLE_SLACK", "false").lower()
+    if enable_slack_raw not in {"true", "false"}:
+        logger.error(
+            f"Invalid value for ENABLE_SLACK: '{enable_slack_raw}'. Expected 'true' or 'false'. Deployment aborted."
+        )
+        raise ValueError(
+            f"Invalid ENABLE_SLACK value: '{enable_slack_raw}'. Deployment aborted."
+        )
+
+    enable_slack = enable_slack_raw == "true"
+    if enable_slack:
+        try:
+            app_config["secrets"] = [modal.Secret.from_name(SECRET_NAME)]
+            logger.info(f"Added secret {SECRET_NAME} to Modal app")
+        except Exception as e:
+            logger.warning(f"Could not add secret {SECRET_NAME}: {e}")
+            logger.info(
+                "Continuing without secrets - Slack notifications will be disabled"
+            )
+    else:
+        logger.info(
+            "Slack notifications disabled by default - Modal app created without secrets"
+        )
+
     try:
         volume = modal.Volume.from_name(VOLUME_NAME)
         app_config["volumes"] = {"/mnt": volume}
@@ -167,7 +191,17 @@ def _report_incident(incident_data: dict, model_checksum: str) -> dict:
             logger.warning(f"Could not write to local incident log: {e}")
 
         # 2. Direct Slack notification for high/critical severity (not using ZenML)
-        if incident["severity"] in ("high", "critical"):
+        enable_slack_raw = os.getenv("ENABLE_SLACK", "false").lower()
+        if enable_slack_raw not in {"true", "false"}:
+            logger.error(
+                f"Invalid value for ENABLE_SLACK: '{enable_slack_raw}'. Expected 'true' or 'false'."
+            )
+            # Don't abort incident reporting, just skip Slack notification
+            enable_slack = False
+        else:
+            enable_slack = enable_slack_raw == "true"
+
+        if incident["severity"] in ("high", "critical") and enable_slack:
             try:
                 slack_token = os.getenv("SLACK_BOT_TOKEN")
                 slack_channel = os.getenv("SLACK_CHANNEL_ID", SC.CHANNEL_ID)
@@ -209,6 +243,10 @@ def _report_incident(incident_data: dict, model_checksum: str) -> dict:
                     )
             except Exception as e:
                 logger.warning(f"Failed to send Slack notification: {e}")
+        elif not enable_slack:
+            logger.info(
+                "Slack notifications disabled (use --enable-slack flag to enable)"
+            )
 
         return {
             "status": "reported",

credit-scorer/run.py

@@ -81,25 +81,32 @@
     help="Directory containing configuration files.",
 )
 @click.option(
-    "--auto-approve",
+    "--manual-approve",
     is_flag=True,
     default=False,
-    help="Auto-approve deployment (for CI/CD pipelines).",
+    help="Require manual approval for deployment (disables auto-approve).",
 )
 @click.option(
     "--no-cache",
     is_flag=True,
     default=False,
     help="Disable caching for pipeline runs.",
 )
+@click.option(
+    "--enable-slack",
+    is_flag=True,
+    default=False,
+    help="Enable Slack notifications in deployment (requires Modal secrets setup).",
+)
 def main(
     feature: bool = False,
     train: bool = False,
     deploy: bool = False,
     all: bool = False,
     config_dir: str = "src/configs",
-    auto_approve: bool = True,
+    manual_approve: bool = False,
     no_cache: bool = False,
+    enable_slack: bool = False,
 ):
     """Main entry point for EU AI Act compliance pipelines.
 
@@ -115,19 +122,28 @@ def main(
     if not config_dir.exists():
         raise ValueError(f"Configuration directory {config_dir} not found")
 
-    # Handle auto-approve setting for deployment
+    # Handle approval setting for deployment (auto-approve is now default)
+    auto_approve = not manual_approve
     if auto_approve:
         os.environ["DEPLOY_APPROVAL"] = "y"
         os.environ["APPROVER"] = "automated_ci"
         os.environ["APPROVAL_RATIONALE"] = (
-            "Automatic approval via --auto-approve flag"
+            "Automatic approval (default behavior)"
         )
 
+    # Handle Slack setting for deployment (Slack disabled by default)
+    if enable_slack:
+        os.environ["ENABLE_SLACK"] = "true"
+
     # Common pipeline options
     pipeline_args = {}
     if no_cache:
         pipeline_args["enable_cache"] = False
 
+    # Handle --all flag first
+    if all:
+        feature = train = deploy = True
+
     # Track outputs for chaining pipelines
     outputs = {}
 
@@ -162,15 +178,18 @@ def main(
 
         train_args = {}
 
-        # Use outputs from previous pipeline if available
-        if "train_df" in outputs and "test_df" in outputs:
-            train_args["train_df"] = outputs["train_df"]
-            train_args["test_df"] = outputs["test_df"]
+        # Don't pass DataFrame artifacts directly - let training pipeline fetch them
+        # from artifact store via Client.get_artifact_version() as designed
 
         training_pipeline = training.with_options(**pipeline_args)
-        model, eval_results, eval_visualization, risk_scores, *_ = (
-            training_pipeline(**train_args)
-        )
+        (
+            model,
+            eval_results,
+            eval_visualization,
+            risk_scores,
+            risk_visualization,
+            *_,
+        ) = training_pipeline(**train_args)
 
         # Store for potential chaining
         outputs["model"] = model
@@ -188,21 +207,15 @@ def main(
 
         deploy_args = {}
 
-        if "model" in outputs:
-            deploy_args["model"] = outputs["model"]
-        if "evaluation_results" in outputs:
-            deploy_args["evaluation_results"] = outputs["evaluation_results"]
-        if "risk_scores" in outputs:
-            deploy_args["risk_scores"] = outputs["risk_scores"]
-        if "preprocess_pipeline" in outputs:
-            deploy_args["preprocess_pipeline"] = outputs["preprocess_pipeline"]
+        # Don't pass artifacts directly - let deployment pipeline fetch them
+        # from artifact store via Client.get_artifact_version() as designed
 
         deployment.with_options(**pipeline_args)(**deploy_args)
 
         logger.info("✅ Deployment pipeline completed")
 
     # If no pipeline specified, show help
-    if not any([feature, train, deploy, all]):
+    if not any([feature, train, deploy]):
         ctx = click.get_current_context()
         click.echo(ctx.get_help())
 

credit-scorer/src/constants/annotations.py

@@ -32,9 +32,9 @@ class StrEnum(str, Enum):
 class Pipelines(StrEnum):
     """Pipeline names used in ZenML."""
 
-    FEATURE_ENGINEERING = "feature_engineering"
-    TRAINING = "training"
-    DEPLOYMENT = "deployment"
+    FEATURE_ENGINEERING = "credit_scoring_feature_engineering"
+    TRAINING = "credit_scoring_training"
+    DEPLOYMENT = "credit_scoring_deployment"
 
 
 class Artifacts(StrEnum):
@@ -58,6 +58,7 @@ class Artifacts(StrEnum):
     EVALUATION_RESULTS = "evaluation_results"
     EVAL_VISUALIZATION = "evaluation_visualization"
     RISK_SCORES = "risk_scores"
+    RISK_VISUALIZATION = "risk_visualization"
     FAIRNESS_REPORT = "fairness_report"
     RISK_REGISTER = "risk_register"
 
@@ -69,6 +70,8 @@ class Artifacts(StrEnum):
     INCIDENT_REPORT = "incident_report"
     COMPLIANCE_RECORD = "compliance_record"
     SBOM_ARTIFACT = "sbom_artifact"
+    SBOM_HTML = "sbom_html"
     ANNEX_IV_PATH = "annex_iv_path"
+    ANNEX_IV_HTML = "annex_iv_html"
     RUN_RELEASE_DIR = "run_release_dir"
     COMPLIANCE_DASHBOARD_HTML = "compliance_dashboard_html"

credit-scorer/src/pipelines/deployment.py

@@ -92,7 +92,7 @@ def deployment(
     )
 
     # Generate Software Bill of Materials for Article 15 (Accuracy & Robustness)
-    generate_sbom(
+    sbom_data, sbom_html = generate_sbom(
         deployment_info=deployment_info,
     )
 
@@ -103,10 +103,12 @@ def deployment(
     )
 
     # Generate comprehensive technical documentation (Article 11)
-    documentation_path, run_release_dir = generate_annex_iv_documentation(
-        evaluation_results=evaluation_results,
-        risk_scores=risk_scores,
-        deployment_info=deployment_info,
+    documentation_path, documentation_html, run_release_dir = (
+        generate_annex_iv_documentation(
+            evaluation_results=evaluation_results,
+            risk_scores=risk_scores,
+            deployment_info=deployment_info,
+        )
     )
 
     # Generate compliance dashboard HTML visualization
@@ -118,5 +120,8 @@ def deployment(
         deployment_info,
         monitoring_plan,
         documentation_path,
+        documentation_html,
+        sbom_data,
+        sbom_html,
         compliance_dashboard,
     )

credit-scorer/src/pipelines/training.py

@@ -91,11 +91,17 @@ def training(
     )
 
     # Perform risk assessment based on evaluation results
-    risk_scores = risk_assessment(
+    risk_scores, risk_visualization = risk_assessment(
         evaluation_results=eval_results,
         risk_register_path=risk_register_path,
         approval_thresholds=approval_thresholds,
     )
 
     # Return artifacts to be used by deployment pipeline
-    return model, eval_results, eval_visualization, risk_scores
+    return (
+        model,
+        eval_results,
+        eval_visualization,
+        risk_scores,
+        risk_visualization,
+    )

credit-scorer/src/steps/deployment/approve.py

@@ -4,6 +4,7 @@
 # you may not use this file except in compliance with the License.
 # You may obtain a copy of the License at http://www.apache.org/licenses/LICENSE-2.0
 
+import os
 import time
 from datetime import datetime
 from typing import Annotated, Any, Dict, Tuple
@@ -285,8 +286,17 @@ def send_slack_message(message, blocks, ask_question=False):
                 print("💡 Fix: Use a Bot User OAuth Token (starts with xoxb-)")
             return None
 
+    # Check for auto-approve from environment variables
+    auto_approve = os.environ.get("DEPLOY_APPROVAL", "").lower() == "y"
+    env_approver = os.environ.get("APPROVER", "")
+    env_rationale = os.environ.get("APPROVAL_RATIONALE", "")
+
     # Send initial notification
-    header = "MODEL AUTO-APPROVED" if all_ok else "HUMAN REVIEW REQUIRED"
+    header = (
+        "MODEL AUTO-APPROVED"
+        if all_ok or auto_approve
+        else "HUMAN REVIEW REQUIRED"
+    )
     send_slack_message(header, create_blocks("Model Approval"))
 
     # Determine approval
@@ -296,6 +306,12 @@ def send_slack_message(message, blocks, ask_question=False):
             "automated_system",
             "All criteria met",
         )
+    elif auto_approve:
+        approved, approver, rationale = (
+            True,
+            env_approver or "automated_ci",
+            env_rationale or "Auto-approved via environment variable",
+        )
     else:
         response = send_slack_message(
             f"Override deployment for pipeline '{pipeline_name}'?",