"Another AI hacker? Let us guess β it runs nmap and writes a report."
v1demo.mp4
Prerequisites: Docker and Docker Compose v2. Supported on macOS (Apple Silicon + Intel), Linux (amd64 + arm64), and Windows (amd64 + arm64) β native via PowerShell or via WSL2 (Ubuntu / Kali).
macOS / Linux / WSL2
curl -fsSL https://decepticon.red/install | bash
decepticon onboard # Interactive setup wizard (provider, API key, model profile)
decepticon # Start everything: terminal CLI + web dashboard at http://localhost:3000Windows (PowerShell, native)
irm https://decepticon.red/install.ps1 | iex
decepticon onboard
decepticonβ Quick start Β· Full setup walkthrough
Building on top of the agents β a SaaS, a research integration, or a custom orchestrator? Install the SDK from PyPI:
pip install decepticon # core SDK
pip install "decepticon[neo4j]" # + the knowledge-graph attack-chain toolsdecepticon is a client SDK: it ships the agent factories, middleware, tools, and skills, and routes LLM calls and sandbox execution to runtime services over HTTP (DECEPTICON_LLM__PROXY_URL, SAAS_SANDBOX_URL). Running agents still needs those services β use the Docker stack above, or point the URLs at your own equivalents. See Decepticon as a library for the factory override surface, declarative PluginBundle plugins, and the safety gate.
We're building Decepticon toward an Offensive Vaccine for the AI-driven threat landscape. If you believe in autonomous red teaming as a path to stronger defense, consider supporting the project.
| Benchmark | Difficulty | Pass Rate |
|---|---|---|
| XBOW validation-benchmarks | Easy (Level 1) | 45 / 45 (100 %) |
| XBOW validation-benchmarks | Medium (Level 2) | 50 / 51 (98.0 %) |
| XBOW validation-benchmarks | Hard (Level 3) | 7 / 8 (87.5 %) |
| XBOW validation-benchmarks | All levels | 102 / 104 (98.08 %) |
- Full per-challenge index, attack-class matrix, and LangSmith traces
- Comparison vs other AI pentest agents (Strix, PentestGPT, MAPTA, Cyber-AutoAgent, XBOW commercial, β¦)
The "AI + hacking" space is full of demos that run nmap and print a report. That's not what this is.
Decepticon is a professional autonomous Red Team agent. It executes realistic attack chains β reconnaissance, exploitation, privilege escalation, lateral movement, C2 β the way a real adversary would, not the way a scanner does.
But more importantly: it operates under the discipline that separates red teamers from script kiddies. Before a single packet leaves the wire, Decepticon generates a complete engagement package β RoE, ConOps, Deconfliction Plan, and OPPLAN with MITRE ATT&CK mapping β and every action runs inside those defined rules.
β Engagement workflow deep dive
Real kill chains, not checkbox scans. Decepticon reads an OPPLAN and pursues objectives through whatever path opens up β pivoting, adapting, chaining techniques.
Interactive shells, actually. Real offensive tools are interactive (msfconsole, sliver-client, evil-winrm). Decepticon runs every command inside persistent tmux sessions with automatic prompt detection β so when a tool drops into an interactive prompt, the agent sends follow-up commands without workarounds.
Hardened sandbox isolation. All commands run inside a Kali Linux sandbox on a dedicated operational network (sandbox-net), separate from the management plane (decepticon-net). LangGraph drives the sandbox via the Docker socket. β Architecture
Offense serves defense. The planned Offensive Vaccine loop will turn findings into defense improvements through an attack β defend β verify cycle.
Two-network design β management services (LiteLLM, PostgreSQL, LangGraph, Web) on decepticon-net; sandbox, C2 server, and targets on sandbox-net. Neo4j is dual-homed so the agent (on management) can persist findings written from inside the sandbox.
β Architecture deep dive Β· Knowledge graph
16 specialist agents organized by kill chain phase, with a fresh context window per objective β no accumulated noise.
Orchestration Β· Reconnaissance Β· Exploitation Β· Post-Exploitation Β· Vulnerability Research Β· Domain Specialists (AD, Cloud, Smart Contracts, Reversing, Analyst).
β Full agent roster and middleware stack
Tier-based, credentials-aware fallback chain. You declare which credentials you have in priority order; Decepticon builds the primaryβfallback chain at every tier from there.
| Profile | Tier per agent | Use case |
|---|---|---|
| eco (default) | Per-agent (HIGH for orchestrator/exploiter/patcher/analyst, MID for execution, LOW for recon/soundwave) | Production |
| max | Every agent on HIGH | High-value targets |
| test | Every agent on LOW | Development / CI |
Tier-mapped providers: Anthropic, OpenAI, Google Gemini, MiniMax, DeepSeek, xAI, Mistral, OpenRouter, Nvidia NIM, Ollama (local). Subscription OAuth: Claude Max/Pro/Team, ChatGPT Pro/Plus/Team, Gemini Advanced, Copilot Pro, SuperGrok, Perplexity Pro.
Configure via decepticon onboard. β Full model reference & fallback examples
| Topic | Doc |
|---|---|
| Installation and first engagement | Getting Started |
| Complete setup, OAuth, providers, dashboard | Setup Guide |
| All CLI commands and keyboard shortcuts | CLI Reference |
All make targets |
Makefile Reference |
| Agent roster and middleware | Agents |
| Model profiles and fallback chain | Models |
| Skill system and format spec | Skills |
| Web dashboard features and setup | Web Dashboard |
| System architecture and network isolation | Architecture |
| Neo4j knowledge graph | Knowledge Graph |
| End-to-end engagement workflow | Engagement Workflow |
| Offensive Vaccine loop | Offensive Vaccine |
| Contributing to Decepticon | Contributing |
git clone https://github.com/PurpleAILAB/Decepticon.git
cd Decepticon
make dogfood # Full OSS UX (launcher β onboard β CLI) on local code
make dev # Backend hot-reload (compose watch) β daily dev loopJoin the Discord β ask questions, share engagement logs, discuss techniques.
Do not use this project on any system or network without explicit written authorization from the system owner. Unauthorized access to computer systems is illegal. You are solely responsible for your actions. The authors and contributors assume no liability for misuse.
