Spaces:
Sleeping
Sleeping
Upload 3 files
Browse files- README.md +62 -62
- app.py +64 -65
- requirements.txt +2 -2
README.md
CHANGED
|
@@ -1,62 +1,62 @@
|
|
| 1 |
-
---
|
| 2 |
-
title: SentiTrust AI Security Toolkit
|
| 3 |
-
emoji: 🛡️
|
| 4 |
-
colorFrom: red
|
| 5 |
-
colorTo: gray
|
| 6 |
-
sdk: gradio
|
| 7 |
-
sdk_version: 5.49.1
|
| 8 |
-
app_file: app.py
|
| 9 |
-
pinned: false
|
| 10 |
-
---
|
| 11 |
-
|
| 12 |
-
# SentiTrust AI Red-Teaming Toolkitst AI Security Toolkit
|
| 13 |
-
emoji: 🛡️
|
| 14 |
-
colorFrom: red
|
| 15 |
-
colorTo: gray
|
| 16 |
-
sdk: gradio
|
| 17 |
-
sdk_version: 4.19.0
|
| 18 |
-
app_file: app.py
|
| 19 |
-
pinned: false
|
| 20 |
-
---: SentiTrust AI Security Toolkit
|
| 21 |
-
emoji: �
|
| 22 |
-
colorFrom: red
|
| 23 |
-
colorTo: gray
|
| 24 |
-
sdk: gradio
|
| 25 |
-
sdk_version: 4.19.0
|
| 26 |
-
app_file: app.py
|
| 27 |
-
pinned: false
|
| 28 |
-
---
|
| 29 |
-
|
| 30 |
-
# SentiTrust AI Red-Teaming Toolkit
|
| 31 |
-
|
| 32 |
-
Professional AI security testing platform for enterprise applications.
|
| 33 |
-
|
| 34 |
-
## Capabilities
|
| 35 |
-
|
| 36 |
-
**Framework-Aligned Testing**
|
| 37 |
-
- Jailbreak vulnerability detection using industry-standard attack patterns
|
| 38 |
-
- Prompt injection analysis with OWASP LLM Top 10 mapping
|
| 39 |
-
- Automated vulnerability classification (MITRE ATLAS, NIST AI RMF)
|
| 40 |
-
- Professional risk reports with CVSS scoring
|
| 41 |
-
|
| 42 |
-
**Attack Techniques**
|
| 43 |
-
- 9 jailbreak methods (DAN, AIM, role-play, system leak, constraint breaking)
|
| 44 |
-
- 8 injection patterns (direct, indirect, context switching, delimiter escape)
|
| 45 |
-
- Multi-stage attack simulation
|
| 46 |
-
- Evidence tracking and reproducible testing
|
| 47 |
-
|
| 48 |
-
## Professional Services
|
| 49 |
-
|
| 50 |
-
Need expert security assessments for your AI application?
|
| 51 |
-
|
| 52 |
-
**Express Audit** ($2,500) - 3-day rapid assessment
|
| 53 |
-
**Professional Assessment** ($8,500) - Comprehensive 2-week security review
|
| 54 |
-
**Enterprise Red Team** ($25,000+) - Full adversarial testing program
|
| 55 |
-
|
| 56 |
-
■ [Visit SentiTrust.com](https://sentitrust.adogent.com)
|
| 57 |
-
■ [Book Free 15-Min Security Scan](https://calendly.com/sentitrust/free-security-scan)
|
| 58 |
-
|
| 59 |
-
---
|
| 60 |
-
|
| 61 |
-
**Built by SentiTrust** - Professional AI security assessments for startups and enterprises.
|
| 62 |
-
Trusted by 50+ AI companies • Prevented $20M+ in potential breaches
|
|
|
|
| 1 |
+
---
|
| 2 |
+
title: SentiTrust AI Security Toolkit
|
| 3 |
+
emoji: 🛡️
|
| 4 |
+
colorFrom: red
|
| 5 |
+
colorTo: gray
|
| 6 |
+
sdk: gradio
|
| 7 |
+
sdk_version: 5.49.1
|
| 8 |
+
app_file: app.py
|
| 9 |
+
pinned: false
|
| 10 |
+
---
|
| 11 |
+
|
| 12 |
+
# SentiTrust AI Red-Teaming Toolkitst AI Security Toolkit
|
| 13 |
+
emoji: 🛡️
|
| 14 |
+
colorFrom: red
|
| 15 |
+
colorTo: gray
|
| 16 |
+
sdk: gradio
|
| 17 |
+
sdk_version: 4.19.0
|
| 18 |
+
app_file: app.py
|
| 19 |
+
pinned: false
|
| 20 |
+
---: SentiTrust AI Security Toolkit
|
| 21 |
+
emoji: �
|
| 22 |
+
colorFrom: red
|
| 23 |
+
colorTo: gray
|
| 24 |
+
sdk: gradio
|
| 25 |
+
sdk_version: 4.19.0
|
| 26 |
+
app_file: app.py
|
| 27 |
+
pinned: false
|
| 28 |
+
---
|
| 29 |
+
|
| 30 |
+
# SentiTrust AI Red-Teaming Toolkit
|
| 31 |
+
|
| 32 |
+
Professional AI security testing platform for enterprise applications.
|
| 33 |
+
|
| 34 |
+
## Capabilities
|
| 35 |
+
|
| 36 |
+
**Framework-Aligned Testing**
|
| 37 |
+
- Jailbreak vulnerability detection using industry-standard attack patterns
|
| 38 |
+
- Prompt injection analysis with OWASP LLM Top 10 mapping
|
| 39 |
+
- Automated vulnerability classification (MITRE ATLAS, NIST AI RMF)
|
| 40 |
+
- Professional risk reports with CVSS scoring
|
| 41 |
+
|
| 42 |
+
**Attack Techniques**
|
| 43 |
+
- 9 jailbreak methods (DAN, AIM, role-play, system leak, constraint breaking)
|
| 44 |
+
- 8 injection patterns (direct, indirect, context switching, delimiter escape)
|
| 45 |
+
- Multi-stage attack simulation
|
| 46 |
+
- Evidence tracking and reproducible testing
|
| 47 |
+
|
| 48 |
+
## Professional Services
|
| 49 |
+
|
| 50 |
+
Need expert security assessments for your AI application?
|
| 51 |
+
|
| 52 |
+
**Express Audit** ($2,500) - 3-day rapid assessment
|
| 53 |
+
**Professional Assessment** ($8,500) - Comprehensive 2-week security review
|
| 54 |
+
**Enterprise Red Team** ($25,000+) - Full adversarial testing program
|
| 55 |
+
|
| 56 |
+
■ [Visit SentiTrust.com](https://sentitrust.adogent.com)
|
| 57 |
+
■ [Book Free 15-Min Security Scan](https://calendly.com/sentitrust/free-security-scan)
|
| 58 |
+
|
| 59 |
+
---
|
| 60 |
+
|
| 61 |
+
**Built by SentiTrust** - Professional AI security assessments for startups and enterprises.
|
| 62 |
+
Trusted by 50+ AI companies • Prevented $20M+ in potential breaches
|
app.py
CHANGED
|
@@ -4,71 +4,70 @@ from pathlib import Path
|
|
| 4 |
from huggingface_hub import snapshot_download
|
| 5 |
|
| 6 |
PRIVATE_REPO = "cybercentinel/ai-red-teaming"
|
| 7 |
-
|
| 8 |
|
| 9 |
-
|
| 10 |
-
|
| 11 |
-
|
| 12 |
-
|
| 13 |
-
|
| 14 |
-
|
| 15 |
-
|
| 16 |
-
|
| 17 |
-
|
| 18 |
-
|
| 19 |
-
|
| 20 |
-
|
| 21 |
-
|
| 22 |
-
|
| 23 |
-
|
| 24 |
-
|
| 25 |
-
|
| 26 |
-
|
| 27 |
-
|
| 28 |
-
|
| 29 |
-
|
| 30 |
-
# The app should auto-launch when imported since it has demo.launch() at module level
|
| 31 |
-
|
| 32 |
-
except Exception as e:
|
| 33 |
-
# Show professional error page if loading fails
|
| 34 |
-
print(f"Error loading toolkit: {e}")
|
| 35 |
-
import gradio as gr
|
| 36 |
|
| 37 |
-
|
| 38 |
-
|
| 39 |
-
|
| 40 |
-
|
| 41 |
-
|
| 42 |
-
|
| 43 |
-
|
| 44 |
-
|
| 45 |
-
|
| 46 |
-
|
| 47 |
-
|
| 48 |
-
|
| 49 |
-
|
| 50 |
-
|
| 51 |
-
|
| 52 |
-
|
| 53 |
-
|
| 54 |
-
|
| 55 |
-
|
| 56 |
-
|
| 57 |
-
|
| 58 |
-
|
| 59 |
-
|
| 60 |
-
|
| 61 |
-
|
| 62 |
-
|
| 63 |
-
</
|
|
|
|
|
|
|
|
|
|
|
|
|
| 64 |
</div>
|
| 65 |
-
"
|
| 66 |
-
|
| 67 |
-
|
| 68 |
-
|
| 69 |
-
|
| 70 |
-
|
| 71 |
-
|
| 72 |
-
|
| 73 |
-
|
| 74 |
-
|
|
|
|
|
|
| 4 |
from huggingface_hub import snapshot_download
|
| 5 |
|
| 6 |
PRIVATE_REPO = "cybercentinel/ai-red-teaming"
|
| 7 |
+
CACHE_DIR = Path("/tmp/sentitrust_cache")
|
| 8 |
|
| 9 |
+
# Download the private space
|
| 10 |
+
print("Loading SentiTrust AI Security Toolkit...")
|
| 11 |
+
CACHE_DIR.mkdir(parents=True, exist_ok=True)
|
| 12 |
+
|
| 13 |
+
try:
|
| 14 |
+
snapshot_download(
|
| 15 |
+
repo_id=PRIVATE_REPO,
|
| 16 |
+
repo_type="space",
|
| 17 |
+
local_dir=CACHE_DIR,
|
| 18 |
+
token=os.environ.get("HF_TOKEN")
|
| 19 |
+
)
|
| 20 |
+
|
| 21 |
+
print("Starting AI Red-Teaming Toolkit...")
|
| 22 |
+
|
| 23 |
+
# Add to path and execute the downloaded app
|
| 24 |
+
sys.path.insert(0, str(CACHE_DIR))
|
| 25 |
+
|
| 26 |
+
# Execute the app.py file directly
|
| 27 |
+
app_path = CACHE_DIR / "app.py"
|
| 28 |
+
with open(app_path) as f:
|
| 29 |
+
exec(f.read(), {"__name__": "__main__"})
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 30 |
|
| 31 |
+
except Exception as e:
|
| 32 |
+
# Show professional error page if loading fails
|
| 33 |
+
print(f"Error loading toolkit: {e}")
|
| 34 |
+
import traceback
|
| 35 |
+
traceback.print_exc()
|
| 36 |
+
|
| 37 |
+
import gradio as gr
|
| 38 |
+
|
| 39 |
+
with gr.Blocks() as demo:
|
| 40 |
+
gr.HTML("""
|
| 41 |
+
<div style="background: #0a0a0a; padding: 60px 30px; border-radius: 12px; text-align: center; color: white; max-width: 800px; margin: 40px auto;">
|
| 42 |
+
<h1 style="color: #dc2626; font-size: 2.5em; margin-bottom: 20px;">
|
| 43 |
+
SentiTrust AI Security Toolkit
|
| 44 |
+
</h1>
|
| 45 |
+
<p style="color: rgba(255,255,255,0.8); font-size: 1.2em; margin-bottom: 30px;">
|
| 46 |
+
The toolkit is temporarily unavailable. Please try again in a moment.
|
| 47 |
+
</p>
|
| 48 |
+
<p style="color: rgba(255,255,255,0.6); font-size: 0.95em; margin-bottom: 40px;">
|
| 49 |
+
If the issue persists, please contact our team.
|
| 50 |
+
</p>
|
| 51 |
+
<div style="margin-top: 30px;">
|
| 52 |
+
<a href="https://sentitrust.adogent.com" target="_blank"
|
| 53 |
+
style="display: inline-block; background: #dc2626; color: white; padding: 16px 32px;
|
| 54 |
+
border-radius: 8px; text-decoration: none; font-weight: 700; margin: 8px; font-size: 1.1em;">
|
| 55 |
+
Visit SentiTrust.com →
|
| 56 |
+
</a>
|
| 57 |
+
<a href="https://calendly.com/sentitrust/free-security-scan" target="_blank"
|
| 58 |
+
style="display: inline-block; background: white; color: #0a0a0a; padding: 16px 32px;
|
| 59 |
+
border-radius: 8px; text-decoration: none; font-weight: 700; margin: 8px; font-size: 1.1em;">
|
| 60 |
+
Book Free Scan →
|
| 61 |
+
</a>
|
| 62 |
</div>
|
| 63 |
+
<p style="color: rgba(255,255,255,0.4); font-size: 0.85em; margin-top: 40px;">
|
| 64 |
+
hello@adogent.com
|
| 65 |
+
</p>
|
| 66 |
+
</div>
|
| 67 |
+
""")
|
| 68 |
+
|
| 69 |
+
demo.launch(
|
| 70 |
+
server_name="0.0.0.0",
|
| 71 |
+
server_port=7860,
|
| 72 |
+
share=False
|
| 73 |
+
)
|
requirements.txt
CHANGED
|
@@ -1,2 +1,2 @@
|
|
| 1 |
-
gradio==
|
| 2 |
-
huggingface-hub
|
|
|
|
| 1 |
+
gradio==5.49.1
|
| 2 |
+
huggingface-hub>=0.30.0
|