AI Features Walkthrough¶

CISentinel integrates AI-powered analysis to provide actionable insights from scan results.
Supported Providers¶
| Provider | Model Example |
|---|---|
| Google Gemini | gemini/gemini-2.5-flash |
| OpenAI | openai/gpt-4o |
| Anthropic | anthropic/claude-3.5-sonnet |
| Mistral | mistral/mistral-large |
| Groq | groq/llama-3.1-70b |
| Ollama (Local) | ollama/llama3.1 |
Checking AI Status¶
┏━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━┓
┃ Setting ┃ Value ┃
┡━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━┩
│ Provider │ gemini │
│ Model │ gemini/gemini-2.5-flash │
│ API Key │ Configured │
│ API Base │ default │
│ Env GEMINI_API_KEY │ Set │
└─────────────────────┴──────────────────────────┘
Configuring AI¶
Via CLI flags¶
cis-hardening-tool ai config \
--provider gemini \
--model gemini/gemini-2.5-flash \
--api-key YOUR_API_KEY_HERE
Via interactive prompt (hidden input)¶
Via environment variable¶
Settings are persisted in the SQLite database and survive restarts.
Generating AI Summary¶
For a specific scan¶
For the latest scan¶
Force regeneration (bypass cache)¶
Auto-Summary After Scan¶
When an API key is configured, CISentinel automatically generates an AI summary after every interactive scan:
Generating AI Summary...
╭─────────── AI Insights ───────────╮
│ │
│ ## Security Assessment Summary │
│ │
│ **Overall Risk Level: Medium** │
│ │
│ ### Critical Findings (3): │
│ 1. SSH root login enabled │
│ 2. Firewall not configured │
│ 3. Password complexity weak │
│ │
│ ### Recommendations: │
│ - Disable PermitRootLogin in │
│ /etc/ssh/sshd_config │
│ - Enable ufw/iptables firewall │
│ - Set password complexity via PAM │
│ │
╰────────────────────────────────────╯
Caching¶
AI summaries are cached in the database. Subsequent requests for the same scan return the cached result unless --force is used.
Web GUI AI Configuration¶
The web dashboard provides an AI settings modal accessible via the ✨ sparkles button in the navbar. Settings configured via CLI are reflected in the GUI and vice versa.