Hostile Command Suite
MCP servers for automated penetration testing and OSINT.
Installation
npx hostile-command-suiteAsk AI about Hostile Command Suite
Powered by Claude ยท Grounded in docs
I know everything about Hostile Command Suite. Ask me about installation, configuration, usage, or troubleshooting.
0/500
Reviews
Documentation
Hostile Command Suite - OSINT Package
Author: cycloarcane
Contact: cycloarkane@gmail.com
License: PolyForm Noncommercial License 1.0.0
Intelligent Open Source Intelligence Investigation System
A terminal-based OSINT investigation framework with AI-powered analysis and intelligent agent decision-making. Features automated profile scraping, multi-platform username investigation, and local LLM integration for enhanced intelligence gathering.
๐ Quick Start
Prerequisites
- Install ollama (for AI analysis):
curl -fsSL https://ollama.ai/install.sh | sh
ollama pull qwen3:8b # recommended model
- Install OSINT tools:
# Arch Linux
yay -S sherlock-git mosint
# Ubuntu/Debian
pip install sherlock-project
# For mosint, download from: https://github.com/alpkeskin/mosint
Installation
git clone https://github.com/cycloarcane/Hostile-Command-Suite.git
cd Hostile-Command-Suite
python -m venv .venv
source .venv/bin/activate # Linux/Mac
pip install -r requirements.txt
Usage
Interactive comprehensive investigation:
python3 HCSO.py --interactive
# Then provide ALL target information: names, usernames, emails, addresses, etc.
Command line investigation:
# Single targets (backward compatibility)
python3 HCSO.py cycloarcane
python3 HCSO.py user@example.com
# Comprehensive targets (multiple data points)
python3 HCSO.py "John Smith, @johnsmith123, john@example.com, works at Acme Corp"
python3 HCSO.py --model llama3.2 "Jane Doe jane.doe@company.com https://linkedin.com/in/janedoe"
๐ ๏ธ Features
Core Capabilities
- ๐ Username Investigation: Sherlock integration across 400+ social media platforms
- ๐ง Email Investigation: Mosint integration for email intelligence and breach analysis
- ๐ Profile Scraping: Automated extraction of profile details from discovered accounts
- ๐ Web Search Intelligence: DuckDuckGo search integration for comprehensive OSINT gathering
- ๐ค AI Agent: Local ollama integration for intelligent decision-making and analysis
- โก Intelligent Workflow: Automatic tool chaining and investigation pivoting
- ๐จ Rich Terminal: Professional red/black themed interface with progress indicators
Supported Targets
| Target Type | Primary Tool | Secondary Tools | AI Analysis |
|---|---|---|---|
| Username | Sherlock โ Profile Scraper | DuckDuckGo Search, Link Analyzer | โ Full Analysis |
| Mosint | DuckDuckGo Search | โ Full Analysis | |
| Any Target | DuckDuckGo Search | Context-dependent pivoting | โ Full Analysis |
MCP Tool Architecture
The system uses Model Context Protocol (MCP) based tool servers:
sherlock_server.py: Username investigation across platformsmosint_server.py: Email enumeration and breach investigationprofile_scraper_server.py: Intelligent profile content extractionduckduckgo_server.py: Web search for comprehensive intelligence gatheringlink_analyzer_server.py: Deep analysis of URLs and GitHub profiles
AI-Powered Intelligence
The AI agent provides:
- Comprehensive Data Extraction: Uses LLM to parse and categorize all provided target information
- Intelligent Tool Selection: Automatically chooses appropriate tools based on data types:
- Names โ DuckDuckGo web search for public records and news
- Usernames โ Sherlock for social media platform discovery
- Emails โ Mosint for breach data and domain analysis
- Organizations โ Web search for corporate intelligence
- URLs โ Link analyzer for deep content analysis
- Investigation Pivoting: Discovers new leads and suggests follow-up actions
- Security Risk Assessment: Evaluates exposure levels and security implications
- Pattern Recognition: Identifies connections across platforms and data sources
- Decision Making: Determines when investigations are complete vs need continuation
๐๏ธ Architecture
Intelligent Agent Design
โโโโโโโโโโโโโโโโโโโ โโโโโโโโโโโโโโโโโโโโ โโโโโโโโโโโโโโโโโโโ
โ HCSO Agent โโโโถโ Ollama AI โโโโโถโ Investigation โ
โ โ โ Decision Engine โ โ Recommendations โ
โโโโโโโโโโโโโโโโโโโ โโโโโโโโโโโโโโโโโโโโ โโโโโโโโโโโโโโโโโโโ
โ
โผ
โโโโโโโโโโโโโโโโโโโ โโโโโโโโโโโโโโโโโโโโ โโโโโโโโโโโโโโโโโโโ
โ MCP Tool โโโโถโ Tool Results โโโโโถโ Profile Scraper โ
โ Manager โ โ Analysis โ โ Auto-Trigger โ
โโโโโโโโโโโโโโโโโโโ โโโโโโโโโโโโโโโโโโโโ โโโโโโโโโโโโโโโโโโโ
โ
โผ
โโโโโโโโโโโโโโโโโโโ โโโโโโโโโโโโโโโโโโโโ โโโโโโโโโโโโโโโโโโโ
โ Sherlock โ โ Mosint โ โ Profile โ
โ Username Search โ โ Email Intel โ โ Scraper โ
โโโโโโโโโโโโโโโโโโโ โโโโโโโโโโโโโโโโโโโโ โโโโโโโโโโโโโโโโโโโ
Investigation Workflow
- Comprehensive Input: User provides ALL available target information (names, usernames, emails, addresses, organizations, URLs, etc.)
- AI Data Extraction: LLM parses and categorizes information into structured data types
- Intelligent Tool Selection: System automatically selects appropriate tools for each data type:
- Names โ Web search for public intelligence
- Usernames โ Social media platform discovery
- Emails โ Breach analysis and domain intelligence
- Organizations โ Corporate and public records search
- URLs โ Deep content and profile analysis
- Parallel Investigation: Multiple tools execute simultaneously based on extracted data
- AI Analysis: Intelligent analysis of all findings and cross-reference discovery
- Decision Point: AI recommends additional investigations or marks complete
- Iterative Enhancement: Follow-up investigations based on discovered leads
๐ Example Output
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ
โ โโ โโ โโโโโโโ โโโโโโ โโโโโโโ โโโโโโโ โโ โโโ โโ โโโโโโโโ โ
โ โโ โโ โโ โโ โโ โโ โโ โโ โโโโ โโ โโ โ
โ โโโโโโโ โโโโโโโ โโ โโโโโ โโ โโ โโโโโโโ โโ โโ โโ โโ โโ โ
โ โโ โโ โโ โโ โโ โโ โโ โโ โโ โโ โโ โโ โ
โ โโ โโ โโโโโโโ โโโโโโ โโโโโโโ โโโโโโโ โโ โโ โโโโ โโ โ
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ
Hostile Command Suite - OSINT Package
Intelligent Open Source Intelligence Investigation System
Using AI Model: qwen3:8b
Available Tools: sherlock, mosint, profile_scraper, duckduckgo_search, link_analyzer
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ
โโโ COMPREHENSIVE TARGET INFORMATION โโโ
Provide ALL available information about your target for intelligent analysis
Include: names, usernames, emails, addresses, organizations, social profiles, etc.
Enter ALL target information: John Smith, @johnsmith123, john.smith@techcorp.com, works at TechCorp
Analyzing Provided Information
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ
Extracted Target Intelligence
โโโโโโโโโโโโโโโโโโณโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ
โ Data Type โ Extracted Values โ
โกโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฉ
โ Names โ John Smith โ
โ Usernames โ johnsmith123 โ
โ Emails โ john.smith@techcorp.com โ
โ Organizations โ TechCorp โ
โโโโโโโโโโโโโโโโโโดโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ
Investigating Name: John Smith
SHERLOCK Investigation Results
โโโโโโโโโโโโโโโโโโณโโโโโโโโโโโโโโ
โ Metric โ Value โ
โกโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฉ
โ Target โ cycloarcane โ
โ Accounts Found โ 17 โ
โ Status โ Success โ
โโโโโโโโโโโโโโโโโโดโโโโโโโโโโโโโโ
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ
Found 17 profiles, scraping for additional intelligence...
PROFILE_SCRAPER Investigation Results
โโโโโโโโโโโโโโโโโโโโณโโโโโโโโโโ
โ Metric โ Value โ
โกโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฉ
โ Total Scraped โ 5 โ
โ Successful โ 4 โ
โ With Useful Info โ 4 โ
โ Status โ Success โ
โโโโโโโโโโโโโโโโโโโโดโโโโโโโโโโ
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ
AI Agent Analyzing...
โญโ AI Investigation Analysis โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฎ
โ ANALYSIS: Investigation revealed GitHub profile with security research โ
โ interests (LLMs + red team). High-value intelligence gathered from โ
โ multiple platforms. Profile scraping provided sufficient context. โ
โ โ
โ RECOMMENDATION: Investigation complete - sufficient intelligence gathered โ
โ TOOL: NONE โ
โ TARGET: N/A โ
โ REASONING: Profile analysis reveals technical expertise and security โ
โ focus. No additional tools needed for current investigation scope. โ
โฐโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฏ
๐ก๏ธ Security & Ethics
Responsible Use
- Legal Compliance: All investigations must comply with applicable laws
- Authorization: Only investigate targets you have permission to research
- Rate Limiting: Respects platform rate limits and implements delays
- Local Processing: All AI analysis happens locally via ollama (no cloud)
Privacy & Security
- No Data Persistence: Investigation results are not stored long-term
- Local LLM: AI analysis never leaves your machine
- Tool Isolation: Each OSINT tool runs independently
- Professional Focus: Designed for defensive security and legitimate research
๐ง Configuration
AI Model Selection
# Use different ollama models
python3 HCSO.py --model llama3.2 target
python3 HCSO.py --model qwen3:8b target
python3 HCSO.py --model mixtral target
Configurable Prompts
Agent behavior is configurable via YAML files:
prompts/agent_system.yaml: Core agent instructions and tool selection logicprompts/tool_prompts.yaml: Tool-specific analysis templates
Tool Capabilities
| Tool | Input | Capabilities | Auto-Trigger |
|---|---|---|---|
| DuckDuckGo Search | Names, Organizations | Web intelligence, news, public records | Auto for names |
| Sherlock | Username | 400+ platform search | Auto for usernames |
| Mosint | Breach data, domain intel | Auto for emails | |
| Link Analyzer | URLs | GitHub profiles, web content analysis | Auto for URLs |
| Profile Scraper | URLs | Bio, followers, verification | After Sherlock |
๐ง Development
Adding New OSINT Tools
- Create MCP server in
mcp_tools/new_tool_server.py - Add tool detection in
MCPToolManager.check_available_tools() - Implement tool calling in
MCPToolManager.call_tool() - Add result display in
display_investigation_result() - Update agent prompts for tool selection logic
Architecture Benefits
Previous Complex Architecture:
- 15+ microservices with FastMCP
- PostgreSQL database requirement
- Web UI and API complexity
- Multiple authentication layers
Current Intelligent Agent Architecture:
- Single intelligent agent with MCP tools
- No database required
- Pure terminal interface with AI
- Local ollama for decision-making
- Automatic tool chaining and pivoting
๐ Roadmap
Phase 1: Core Intelligence โ
- Sherlock username investigation with AI analysis
- Mosint email investigation with AI analysis
- Intelligent profile scraping from social media
- AI-powered investigation decision making
- MCP-based tool architecture
Phase 2: Enhanced Analysis ๐ง
- Link analyzer for deep GitHub/social media analysis
- Domain investigation capabilities
- Phone number OSINT integration
- Correlation analysis across findings
Phase 3: Advanced Intelligence ๐
- Investigation session management
- Custom tool integration framework
- Automated investigation workflows
- Advanced AI reasoning and pivoting
๐ค Contributing
- Fork the repository
- Create a feature branch (
git checkout -b feature/new-tool) - Add your MCP tool server following existing patterns
- Update agent prompts for tool integration
- Test with various target types
- Submit a pull request
Development Setup
git clone https://github.com/cycloarcane/Hostile-Command-Suite.git
cd Hostile-Command-Suite
python -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt
python3 HCSO.py --interactive
๐ License
This project is licensed under the PolyForm Noncommercial License 1.0.0 - see the LICENSE file for details.
Copyright ยฉ cycloarcane (cycloarkane@gmail.com)
Intelligent. Terminal. Effective.
Advanced OSINT investigation with AI-powered decision making
For questions or feature requests, contact cycloarkane@gmail.com
