Skip to content

Latest commit

 

History

31 Commits

Folders and files

NameName
Last commit message
Last commit date
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

🛡️⚡ Agent-eBPF: AI Sentinel in Kernel Space

Agent-eBPF is a sub-microsecond (<35µs) Linux kernel space (Ring 0) security shield and telemetry gateway engineered for autonomous AI agents, LLM services, and containerized application swarms.

Operating under Zero-Trust principles, this architecture intercepts destructive database queries, illegal network packets, and unauthorized system calls (such as unconstrained DELETE/UPDATE operations without WHERE clauses) directly inside the Linux kernel (XDP/Ring-Buffer) with zero latency impact.

🚀 Quick Start (1-Click Launch)

Launch the Agent-eBPF system and Visual Web Dashboard instantly without manual setup or configuration:

💻 Windows

Double-click the launcher script in the project root:

start.bat

🐧 Linux / 🍎 macOS

Run the shell script in terminal or file manager:

chmod +x start.sh
./start.sh

💡 Automatic Execution: The launcher validates dependencies, installs necessary Python packages, initializes the eBPF & MCP Gateway server, and automatically opens http://localhost:8000 in your default web browser.


🖥️ Visual Web Control Panel (Web UI)

Manage kernel security operations visually without writing code via the Web UI (http://localhost:8000):

┌────────────────────────────────────────────────────────────────────────┐
│ 🛡️⚡ Agent-eBPF | Autonomous Linux Kernel Shield                       │
├────────────────────────────────────────────────────────────────────────┤
│                                                                        │
│  [1. LIVE TELEMETRY]          [2. AST SANDBOX]                         │
│  • Real-time Kernel Logs      • Test sample SQL query:                 │
│  • Dropped Packet Stream      • UPDATE users SET role='admin'          │
│  • Latency (<35µs)            • [⚡ EVALUATE] -> Blocked (DROP)        │
│                                                                        │
│  [3. RULE POLICIES]           [4. MCP SSE CONNECTION]                  │
│  • policy.yaml Preview        • SSE Endpoint Address:                  │
│  • 1-Click Policy Injection   • http://localhost:8000/sse              │
└────────────────────────────────────────────────────────────────────────┘
  1. Interactive AST Sandbox: Test destructive DELETE or UPDATE queries live and observe kernel-level sub-microsecond interception in real time.
  2. Live Telemetry Stream: Monitor database mutations and system calls issued by AI agents through a streaming event terminal.
  3. Policy & Rule Management: View active security policies (policy.yaml) and dynamically inject enforcement rules.

🤖 AI Agent Integration (MCP SSE)

Agent-eBPF provides native support for Model Context Protocol (MCP) via Server-Sent Events (SSE).

MCP Server Connection

  • SSE Endpoint: http://localhost:8000/sse
  • Messages Endpoint: http://localhost:8000/messages
  • OAuth 2.0 Discovery: http://localhost:8000/.well-known/oauth-authorization-server

Exposed MCP Tools

  • get_security_status: Fetches active kernel hooks, inspection latency, and blocked threat metrics.
  • get_ebpf_status: Retrieves real-time BPF map packet counters and kernel hook state.
  • get_active_policies: Fetches declarative rules loaded from policy.yaml.
  • add_security_rule: Injects new IP block entries or query enforcement rules into kernel memory.
  • simulate_query_check: Validates proposed SQL payloads against active eBPF policies prior to execution.

💻 CLI Command Line Interface

Manage Agent-eBPF via the rich command-line tool cli.py:

# Build eBPF C bytecode
python cli.py build

# Load eBPF program into kernel (eth0)
python cli.py load --iface eth0

# Inspect kernel status and packet counters
python cli.py status

# Stream live RingBuffer security violations
python cli.py events

# Add IP block rule to BPF map
python cli.py add-rule 192.168.1.105 --rule-id 201

# Unload eBPF program from kernel
python cli.py unload --iface eth0

📚 Developer Guide & Technical Specs

For in-depth kernel architecture details, zero-overhead execution specs, declarative rule schemas, and production deployment patterns, refer to the Developer Guide.


🔬 Automated Testing

Run the full pytest suite to verify MCP endpoints, JWT authentication, and eBPF loader contracts:

python -m pytest -v

📜 License

Distributed under the MIT License. Created by Sysauto & Agent-eBPF Core Engineering.

About

Zero-Code, Zero-Overhead Kernel Firewall for AI Workflows & LLM Agents (eBPF / MCP)

Topics

Resources

Contributing

Security policy

Stars

1 star

Watchers

0 watching

Forks

Releases

Packages

Used by

Contributors

Languages