Proprietary Product · WRAPT
WRAPT
Latest version: V1.0.3 Stable| V2.0 Test Preview
AI-powered automated penetration testing platform
For enterprise security teams and security practitioners working within an authorized scope, WRAPT uses AI throughout the penetration testing workflow to make testing faster, more thorough and scalable
Make authorized testing more efficient with reusable expertise
WRAPT consolidates methodology, toolchains and AI capability into a unified platform, automating tedious steps so security teams can focus on where judgment matters most
AI Capability
AI assistance, human expertise
AI handles routine steps throughout testing, while engineers assess findings and make decisions
See the attack surface in minutes
Automatically profile fingerprints and the attack surface of authorized assets, providing full context for subsequent testing
Automated test path planning
Reason over context to infer possible test paths and risk points, then give validation priorities
Automated validation, low false positives
Reproduce and validate suspected vulnerabilities automatically, outputting validation results and impact scope
One-click reporting
Generate reports automatically from templates, with validation records and fix recommendations for faster, more consistent delivery
Core Capabilities
Support for the full security testing workflow
From web to intranet, from validation to delivery — one platform for the complete workflow
Multi-agent orchestration
Planning, scheduling, aggregation and validation roles work together, linking reconnaissance, exploitation, post-exploitation and documentation into a unified execution chain
Evidence-gated risk controls
Validate the authenticity and deliverability of aggregated results; unverified conclusions are kept as leads, preventing erroneous conclusions from driving high-risk actions
Cross-platform execution layer
Connect scripts and tools written in Python, Java, Go, PHP and other languages, with consistent parameter handling, result formats and execution context management
Web console for day-to-day operations
Real-time log viewing, project control, manual intervention, library management, session management and result export, built for team use
RAG & self-learning
Combining knowledge bases, skill libraries and a self-learning pipeline to continuously accumulate methodology, skills and high-quality real-world experience
Integrated reporting
Bring together findings, risk ratings, remediation recommendations and exports, reducing manual compilation and inaccuracies
Use Cases
Adapts to a wide range of needs
Support for testing before release, regular assessments and security exercises
Pre-release security testing
Automated regression before release to avoid shipping with vulnerabilities
Periodic security assessment
Regular reviews to track remediation and reduce outstanding risks
Exercise support
Automated testing support for cybersecurity defense and red team/blue team exercises
Compliance & audit
Auditable test records for MLPS and cryptography assessment
System Architecture
An eight-layer architecture for automated penetration testing and operational control
From access to operations, each layer supports execution, governance, data management and reusable security resources
Access Layer
Web UI and API layer handling page access, interface calls and real-time log streaming
Orchestration Layer
Unified management of project lifecycle, async tasks, event broadcasting and stage progression
Agent Layer
Main, Coordinator, Recon, Exploit and other roles work together to support reconnaissance, exploitation, validation and reporting
Execution Layer
Covers CLI, session backend, cross-platform tool management and report generation modules
Knowledge Layer
Composed of RAG, knowledge bases, skill libraries and a self-learning pipeline for knowledge retrieval and capability expansion
Data Layer
Manages projects, events, exports and session state on structured storage, ensuring persistence and traceability
Asset Layer
Unified management of vulnerability, fingerprint, plugin, POC/EXP libraries and custom tool assets
Operations Layer
Manages startup, health checks, environment compatibility and background tasks to support continuous operation
AI Connectivity
Multiple models with failover between providers
Bring model scheduling and governance into the platform to reduce single-point decision risk
Multiple AI models
Supports DeepSeek, MiniMax, GLM, Qwen, ChatGPT, Gemini, Claude, Ollama and Custom AI. It first tries the primary and fallback models within the current vendor, then automatically switches to the next configured vendor on failure
Governance model
A multi-role governance system of planning, scheduling, aggregation and validation advances tasks through reconnaissance, attack, persistence and documentation capabilities, reducing the risk of uncontrolled single-model, single-thread decisions
WRAPT is intended solely for authorized security testing and compliance checks. Please comply with applicable laws and regulations and never use it for unauthorized purposes