HunterX

AI-Assisted Offensive Security Framework
ObserveHypothesizeProbeVerify

HunterX combines intelligent reconnaissance, adaptive vulnerability discovery, payload orchestration, and explainable security reasoning into a single modular platform. No external scanners required.

Release PyPI Tests License Docker Python

Why HunterX?

Traditional scanners rely on payload volume and signature matching. HunterX reasons about what vulnerabilities might exist before probing, then verifies with evidence.

Tool AI / Reasoning Multi-Agent Payload Intelligence Reporting Architecture
HunterX LLM-native (multi-provider) 10 agents, DAG workflows FTS5-indexed, 5-level policy SARIF, HTML, graph, purple team Unified Python framework
Nmap XML C, single-purpose
Nuclei JSON Go, template engine
Metasploit Console Ruby, framework
Amass JSON Go, single-purpose
ffuf JSON Go, single-purpose

Key Features

Everything you need for depth-oriented security assessment in a single platform.

41 Security Skills

Built-in skills covering web, API, cloud, and infrastructure security. Each carries MITRE ATT&CK, OWASP, CWE, and CAPEC metadata.

🤖

Reasoning Engine

18 goal types across vulnerability detection, risk assessment, exploit verification, and remediation planning with multi-call consensus.

Multi-Agent Platform

10 specialized agents collaborate through event/message buses with DAG-based workflows, checkpoint/resume, and isolated memory.

AI Provider Abstraction

Decoupled AI provider layer supporting OpenAI and Ollama with caching, metrics, middleware, retry, and circuit breaker.

📜

Payload Intelligence

SQLite + FTS5-indexed payload repository with 5-level safety policy, 10-family mutation engine, provenance tracking, and feedback loop.

📝

Community Payload Knowledge

Payloads are sourced from the community-maintained PayloadsAllTheThings knowledge base, synchronized and indexed locally with provenance tracking.

📈

Knowledge Graph

Entity-relationship store for findings, targets, payloads, and attack paths. Enables cross-scan correlation and relationship analysis.

🛡

Threat Modeling

STRIDE/LINDDUN categorization, trust boundary mapping, automated threat scenarios, and kill chain analysis.

📄

Enterprise Reporting

JSON, Markdown, SARIF 2.1 (VS Code / CodeQL), HTML, attack graphs, purple team detection rules, and ZIP evidence packages.

🔒

Safety-by-Design

Non-bypassable destructive payload blocklist, WAF detection with auto-abort, configurable rate limiting, and policy-driven execution.

Architecture

Layered design with clear component responsibilities and defined communication interfaces.

                    +---------------------------+
                    |    CLI / API / Docker      |
                    +---------------------------+
                               |
                    +---------------------------+
                    |  Orchestration Engine      |
                    |  Observe → Hypothesize →  |
                    |  Probe → Verify          |
                    +--+--------+--------+------+
                       |        |        |
              +--------+  +----+----+  +--------+
              | Agents   | Reasoning |  Skills  |
              | Platform | Engine    | Registry |
              | 10 agents| 18 goals  | 41 skills|
              | Event/Msg| Planner   | Executor |
              | Bus      | Validator | Policy   |
              | Workflow | Consensus | Cache    |
              +----+-----+-----+----+-----+----+
                   |           |            |
              +----+-----+-----+----+-----+----+
              | Payload  | Knowledge | Threat   |
              | Intel    | Graph     | Modeling |
              | FTS5 idx | Relations | STRIDE   |
              | Mutation | Paths     | Chains   |
              | Feedback | Context   | Trust    |
              +----+-----+-----------+----------+
                   |
              +----+------+
              | AI Provider|
              | OpenAI     |
              | Ollama     |
              | Sessions   |
              | Caching    |
              | Metrics    |
              +----+------+
                   |
              +----+------+
              | Reporter  |
              | JSON / MD |
              | SARIF /   |
              | HTML      |
              +-----------+

Payload Intelligence is backed by the community-maintained PayloadsAllTheThings knowledge base (hunterx payload sync).

Assessment Workflow

The HunterX reasoning pipeline drives every scan through four distinct phases.

  +----------+     +----------+     +----------+     +----------+     +----------+
  |  Target  |     |  Recon   |     |Reasoning |     | Payload  |     |Reporting |
  |  Input   | →  |  Intel   | →  | Engine   | →  |  Intel   | →  |  Output  |
  | URL/     |     | Finger-  |     | Hypothes-|     | Select   |     | JSON/MD  |
  | Domain   |     | printing |     | ize      |     | Mutate   |     | SARIF    |
  +----------+     | WAF Det. |     | Plan     |     | Execute  |     | HTML     |
                   +----------+     | Verify   |     +----------+     | Graph    |
                                    +----------+                      +----------+
                                              \                          /
                                               \  +----------------+   /
                                                → | Verification  | ←
                                                    | Agent         |
                                                    | Confidence    |
                                                    | Scoring       |
                                                    | FP Filtering  |
                                                    +----------------+

Payload selection draws on the locally indexed PayloadsAllTheThings knowledge base with provenance tracking.

Showcase

HunterX in action — from reconnaissance through reporting.

hunterx scan target.com --profile bounty --preset quick
CLI Scan Execution
SARIF 2.1 • HTML • Knowledge Graph
Reporting Formats
41 Skills • 18 Goals • 10 Agents
Architecture Overview

Community

HunterX is open-source and community-driven. Apache 2.0 licensed.

Related Projects

Ahmed Awad Research AnubisX Framework

About

HunterX is built with a focus on reasoning-driven security assessment, safety-by-design, and multi-format reporting (JSON, HTML, SARIF).

👤 Author

Ahmed Awad (NullC0d3) — Cybersecurity Threat Intelligence Analyst and open-source developer. Creator of HunterX, AnubisX Framework, and RabbitHole.

Learn more →

🎯 Mission

Democratize intelligent vulnerability assessment by combining AI-assisted reasoning with comprehensive security testing in a single, extensible open-source platform.

🛰 Responsible Use

HunterX is provided exclusively for authorized security testing. Users are solely responsible for obtaining authorization before scanning any target.

Read the policy →

📖 License

Released under the Apache License, Version 2.0. Free to use, modify, and distribute for any purpose.

View license →

Documentation

Explore the complete documentation, API reference, tutorials, and guides.

Browse Documentation →