🛡️ Sentinel: [HIGH] Fix XSS vulnerability in HTML report generation

.jules/sentinel.md

@@ -0,0 +1,4 @@
+## 2024-06-06 - [HIGH] Fix XSS vulnerability in HTML report generation
+**Vulnerability:** The application manually generated HTML reports using Python f-strings without escaping dynamic or user-controlled variables. This omission allowed potential Cross-Site Scripting (XSS) if properties like `layer_id`, `data_source`, or anomaly/alert descriptions were crafted with malicious payload and rendered in the browser.
+**Learning:** Manual HTML string concatenation lacks automatic template escaping (like Jinja2), so it is heavily reliant on developers remembering to explicitly sanitize data. Furthermore, applying HTML escaping on string variables requires strict ordering when truncation is used, as `html.escape(string[:140])` must truncate *before* escaping to prevent slicing in the middle of a generated HTML entity.
+**Prevention:** Whenever generating raw HTML using string manipulation or f-strings in Python, always import the `html` module and wrap dynamic variables in `html.escape()`. Truncation, capitalization, or formatting on string contents should always precede the escape function to maintain syntactic integrity of the encoded HTML entities.

backend/app/services/report_service.py

@@ -6,6 +6,7 @@
 """
 from __future__ import annotations
 
+import html
 from datetime import datetime
 from typing import Any, Optional
 
@@ -170,15 +171,38 @@ def generate_html_report(
     frames = pipeline_result.get("frames") or []
 
     # Job metadata
-    layer_id = pipeline_result.get("layer_id", "Unknown")
-    data_source = pipeline_result.get("data_source", "Unknown")
-    status = pipeline_result.get("status", "Unknown")
-    bbox = pipeline_result.get("bbox", [])
-    time_start = pipeline_result.get("time_start")
-    time_end = pipeline_result.get("time_end")
-    created_at = pipeline_result.get("created_at")
-    completed_at = pipeline_result.get("completed_at")
-    error_msg = pipeline_result.get("error")
+    # Secure escaping for HTML
+    safe_job_id = html.escape(str(job_id))
+    safe_job_id_12 = html.escape(str(job_id)[:12])
+    safe_job_id_16 = html.escape(str(job_id)[:16])
+
+    raw_layer_id = pipeline_result.get("layer_id", "Unknown")
+    layer_id = html.escape(str(raw_layer_id))
+
+    raw_data_source = pipeline_result.get("data_source", "Unknown")
+    data_source = html.escape(str(raw_data_source))
+
+    raw_status = pipeline_result.get("status", "Unknown")
+    status = str(raw_status)
+    safe_status_upper = html.escape(status.upper())
+
+    raw_bbox = pipeline_result.get("bbox", [])
+    bbox = [html.escape(str(b)) for b in raw_bbox]
+
+    raw_time_start = pipeline_result.get("time_start")
+    time_start = html.escape(str(raw_time_start)) if raw_time_start else None
+
+    raw_time_end = pipeline_result.get("time_end")
+    time_end = html.escape(str(raw_time_end)) if raw_time_end else None
+
+    raw_created_at = pipeline_result.get("created_at")
+    created_at = html.escape(str(raw_created_at)) if raw_created_at else None
+
+    raw_completed_at = pipeline_result.get("completed_at")
+    completed_at = html.escape(str(raw_completed_at)) if raw_completed_at else None
+
+    raw_error_msg = pipeline_result.get("error")
+    error_msg = html.escape(str(raw_error_msg)) if raw_error_msg else None
 
     # Calculate metrics
     n_total = metrics.get("total_frames", len(frames))
@@ -214,30 +238,34 @@ def generate_html_report(
     # Generate tables
     alert_rows = ""
     for a in (alerts or [])[:25]:
+        desc = html.escape(str(a.get('description', '—'))[:140])
+        a_type = html.escape(str(a.get('type', '—')).replace('_', ' ').capitalize())
         alert_rows += f"""
         <tr>
           <td class="font-mono">{a.get('frame_index', '—')}</td>
-          <td>{str(a.get('type', '—')).replace('_', ' ').capitalize()}</td>
+          <td>{a_type}</td>
           <td>{_sev_badge(a.get('severity', 'low'))}</td>
-          <td style="color: #444444;">{a.get('description', '—')[:140]}</td>
+          <td style="color: #444444;">{desc}</td>
         </tr>"""
 
     traj_rows = ""
     for t in (trajectories or [])[:15]:
+        t_id = html.escape(str(t.get('id', '—')))
         traj_rows += f"""
         <tr>
-          <td class="font-mono">{t.get('id', '—')}</td>
+          <td class="font-mono">{t_id}</td>
           <td class="font-mono">{t.get('speed', 0):.5f}</td>
           <td class="font-mono">{t.get('direction_deg', 0):.1f}&deg;</td>
           <td class="font-mono">{t.get('intensity', 0):.4f}</td>
         </tr>"""
 
     issue_rows = ""
     for iss in (consistency_issues or [])[:20]:
+        issue_type = html.escape(str(iss.get('issue', '—')))
         issue_rows += f"""
         <tr>
           <td class="font-mono">{iss.get('frame', '—')}</td>
-          <td>{iss.get('issue', '—')}</td>
+          <td>{issue_type}</td>
           <td>{_sev_badge(iss.get('severity', 'low'))}</td>
           <td class="font-mono">{iss.get('mad_score', '—')}</td>
         </tr>"""
@@ -246,18 +274,19 @@ def generate_html_report(
     model_rows = ""
     for model, count in sorted(frame_stats["by_model"].items(), key=lambda x: x[1], reverse=True):
         pct = (count / max(n_total, 1)) * 100
-        model_rows += f"<tr><td class='font-mono'>{model}</td><td class='font-mono'>{count}</td><td class='font-mono'>{pct:.1f}%</td></tr>"
+        safe_model = html.escape(str(model))
+        model_rows += f"<tr><td class='font-mono'>{safe_model}</td><td class='font-mono'>{count}</td><td class='font-mono'>{pct:.1f}%</td></tr>"
 
     # Confidence distribution
     frame_stats["by_confidence"]
 
     # Build the comprehensive HTML report
-    html = f"""<!DOCTYPE html>
+    html_content = f"""<!DOCTYPE html>
 <html lang="en">
 <head>
   <meta charset="UTF-8" />
   <meta name="viewport" content="width=device-width, initial-scale=1.0"/>
-  <title>AetherGIS Technical Report — {job_id[:12]}</title>
+  <title>AetherGIS Technical Report — {safe_job_id_12}</title>
   <link rel="preconnect" href="https://fonts.googleapis.com">
   <link rel="preconnect" href="https://fonts.gstatic.com" crossorigin>
   <link href="https://fonts.googleapis.com/css2?family=Barlow:wght@400;500;600&family=Barlow+Condensed:wght@400;500;600;700&family=JetBrains+Mono:wght@400;500&display=swap" rel="stylesheet">
@@ -719,11 +748,11 @@ def generate_html_report(
       <p class="report-subtitle">Satellite Imagery Frame Interpolation Pipeline — Technical Documentation</p>
 
       <div class="report-meta">
-        <span><span class="meta-label">Job ID:</span> {job_id}</span>
+        <span><span class="meta-label">Job ID:</span> {safe_job_id}</span>
         <span><span class="meta-label">Layer:</span> {layer_id}</span>
         <span><span class="meta-label">Status:</span> 
           <span class="status-badge {'status-success' if status == 'completed' else 'status-error' if error_msg else 'status-warning'}">
-            <span class="status-dot"></span>{status.upper()}
+            <span class="status-dot"></span>{safe_status_upper}
           </span>
         </span>
         <span><span class="meta-label">Generated:</span> {now}</span>
@@ -788,7 +817,7 @@ def generate_html_report(
         <div class="info-grid">
           <div class="info-row">
             <div class="info-label">Job ID</div>
-            <div class="info-value">{job_id}</div>
+            <div class="info-value">{safe_job_id}</div>
           </div>
           <div class="info-row">
             <div class="info-label">Data Source</div>
@@ -800,7 +829,7 @@ def generate_html_report(
           </div>
           <div class="info-row">
             <div class="info-label">Status</div>
-            <div class="info-value">{status.upper()}</div>
+            <div class="info-value">{safe_status_upper}</div>
           </div>
           <div class="info-row">
             <div class="info-label">Created At</div>
@@ -973,31 +1002,31 @@ def generate_html_report(
         <h3 class="section-subtitle">Video Sequences</h3>
         <ul class="artifact-list">
           <li>
-            <span class="artifact-path">/exports/{job_id}/original.mp4</span>
+            <span class="artifact-path">/exports/{safe_job_id}/original.mp4</span>
             <span class="artifact-desc">Original observed frame sequence (no interpolation)</span>
           </li>
           <li>
-            <span class="artifact-path">/exports/{job_id}/interpolated.mp4</span>
+            <span class="artifact-path">/exports/{safe_job_id}/interpolated.mp4</span>
             <span class="artifact-desc">Full interpolated sequence (observed + AI frames)</span>
           </li>
         </ul>
 
         <h3 class="section-subtitle">Frame Archive</h3>
         <ul class="artifact-list">
           <li>
-            <span class="artifact-path">/exports/{job_id}/frames/frame_*.png</span>
+            <span class="artifact-path">/exports/{safe_job_id}/frames/frame_*.png</span>
             <span class="artifact-desc">Individual frame images ({n_total} frames, PNG format)</span>
           </li>
         </ul>
 
         <h3 class="section-subtitle">Metadata & Documentation</h3>
         <ul class="artifact-list">
           <li>
-            <span class="artifact-path">/exports/{job_id}/metadata.json</span>
+            <span class="artifact-path">/exports/{safe_job_id}/metadata.json</span>
             <span class="artifact-desc">Complete frame metadata with per-frame metrics</span>
           </li>
           <li>
-            <span class="artifact-path">/exports/{job_id}/report.html</span>
+            <span class="artifact-path">/exports/{safe_job_id}/report.html</span>
             <span class="artifact-desc">This technical analysis report</span>
           </li>
         </ul>
@@ -1100,7 +1129,7 @@ def generate_html_report(
         <h3 class="section-subtitle">Traceability Statement</h3>
         <p style="font-size: 11px; color: var(--t3); line-height: 1.6;">
           This report was auto-generated by AetherGIS v2.0 pipeline system. All metrics 
-          are computed from the actual execution artifacts stored at <code>/exports/{job_id}/</code>. 
+          are computed from the actual execution artifacts stored at <code>/exports/{safe_job_id}/</code>.
           Frame-level metadata includes: source timestamp, interpolation model used, 
           PSNR/SSIM scores (for interpolated frames), confidence classification, and gap 
           category. In case of database record loss, results can be fully reconstructed 
@@ -1118,13 +1147,13 @@ def generate_html_report(
           <strong>Primary Source:</strong> NASA GIBS Earthdata API (Global Imagery Browse Services)<br>
           <strong>Interpolation Engine:</strong> AetherGIS v2.0 with RIFE/FILM optical flow models<br>
           <strong>Processing Location:</strong> AetherGIS Analysis Pipeline (Module 15)<br>
-          <strong>Report ID:</strong> RPT-{job_id[:12]}-{datetime.utcnow().strftime('%Y%m%d')}
+          <strong>Report ID:</strong> RPT-{safe_job_id_12}-{datetime.utcnow().strftime('%Y%m%d')}
         </p>
       </div>
 
       <div class="footer-meta">
         <span>AetherGIS Technical Report</span>
-        <span>Job: {job_id[:16]}</span>
+        <span>Job: {safe_job_id_16}</span>
         <span>Generated: {now}</span>
       </div>
     </div>
@@ -1133,4 +1162,4 @@ def generate_html_report(
 </body>
 </html>"""
 
-    return html
+    return html_content

backend/tests/test_report_service.py

@@ -0,0 +1,54 @@
+from datetime import datetime
+from backend.app.services.report_service import generate_html_report
+
+def test_generate_html_report_escapes_xss():
+    """Verify that potentially malicious input is properly escaped in HTML report generation."""
+
+    malicious_payload = "<script>alert('XSS')</script>"
+
+    pipeline_result = {
+        "metrics": {"total_frames": 1, "observed_frames": 1, "interpolated_frames": 0},
+        "frames": [{"model_used": "Test", "confidence_class": "high", "gap_category": "none"}],
+        "layer_id": malicious_payload,
+        "data_source": malicious_payload,
+        "status": malicious_payload,
+        "error": malicious_payload,
+    }
+
+    alerts = [
+        {
+            "frame_index": 1,
+            "type": malicious_payload,
+            "severity": "high",
+            "description": malicious_payload
+        }
+    ]
+
+    consistency_issues = [
+        {
+            "frame": 1,
+            "issue": malicious_payload,
+            "severity": "medium",
+            "mad_score": 10.0
+        }
+    ]
+
+    trajectories = [
+        {
+            "id": malicious_payload,
+            "speed": 10.0,
+            "direction_deg": 90.0,
+            "intensity": 5.0
+        }
+    ]
+
+    html = generate_html_report(
+        job_id=malicious_payload,
+        pipeline_result=pipeline_result,
+        trajectories=trajectories,
+        alerts=alerts,
+        consistency_issues=consistency_issues
+    )
+
+    assert "<script>" not in html
+    assert "&lt;script&gt;alert(&#x27;XSS&#x27;)&lt;/script&gt;" in html