Cybersentry is an advanced, AI-powered cybersecurity agent framework designed to assist security professionals, CTF players, and bug bounty hunters. It leverages Large Language Models (LLMs) to autonomously plan, execute tools, and analyze results within a secure environment.
Update: Now features a "Gemini-style" interactive CLI with live status updates and witty cyberpunk commentary! 🤖✨
- 🤖 AI-Driven Agents: Deploy specialized agents (e.g., CTF Agent, Blue Team, Bug Hunter) to solve complex security challenges.
- 🔧 Tool Integration: Agents can autonomously execute system commands, network scans, and custom scripts.
- 🌐 Multi-Model Support: Powered by LiteLLM, allowing seamless switching between OpenRouter, OpenAI, Anthropic, and local models.
- ⚡ Smart Rate Limiting: Built-in client-side throttling to prevent provider rate-limit blocks.
- ✨ Interactive UI: Beautiful, animated CLI experience powered by
rich, featuring dynamic status updates while the AI "thinks."
- Python 3.12+
- Linux/WSL2 environment (Recommended)
-
Clone the repository:
git clone [https://github.com/yourusername/cybersentry.git](https://github.com/yourusername/cybersentry.git) cd cybersentry -
Set up the Virtual Environment: Using
uv(Recommended if available):uv sync source .venv/bin/activateOr using standard
pip:python3 -m venv .venv source .venv/bin/activate pip install -e .
Cybersentry relies on environment variables to manage API keys and model selection.
Create a .env file or export these variables in your shell. (Example using OpenRouter):
# Your API Key (OpenRouter keys work with the OpenAI client in LiteLLM)
export OPENAI_API_KEY="sk-or-v1-..."
# Set the Base URL for OpenRouter
export OPENAI_API_BASE="[https://openrouter.ai/api/v1](https://openrouter.ai/api/v1)"
# Set Requests Per Minute (Recommended: 40 for free tiers)
export LITELLM_MAX_RPM=40
# Set Tokens Per Minute
export LITELLM_MAX_TPM=80000
# DEfine model of your choice
export CYBERSENTRY_MODEL="openrouter/mistralai/mistral-7b-instruct:free"
# COmmand to run
cybersentry