pentest-ai
Turn Claude Code into your offensive security research assistant.
28 specialized AI subagents for every phase of authorized penetration testing, from scoping to reporting. 6 Tier 2 agents that execute tools directly with your approval. Autonomous exploit chaining. PoC-validated findings. Agentic swarm orchestration. Business logic flaw detection. CI/CD pipeline integration. MITRE ATT&CK mapped. Dual offensive/defensive perspective in every response.
Getting Started | Agents | Examples | Documentation | Changelog | Landing Page
Table of Contents
- What Is This?
- Agents
- Workflow
- pentest-ai vs. Manual Research
- How pentest-ai Is Different
- Use Cases
- Quick Start
- Running Tools in a Container
- How Agent Routing Works
- Examples
- Prerequisites
- FAQ
- Documentation
- Contributing
- Legal
- License
What Is This?
pentest-ai is a collection of 28 Claude Code subagents: specialized AI assistants that activate automatically based on what you're working on. Ask Claude to plan a pentest, and the engagement planner agent takes over. Paste Nmap output, and the recon advisor analyzes it. Need to Kerberoast an AD environment? The AD attacker agent builds and runs the Impacket commands for you. Each agent carries deep domain knowledge in offensive security methodology, MITRE ATT&CK mappings, and industry-standard frameworks.
You don't need to be an expert to use these agents. They communicate at whatever level you need, from explaining what Kerberoasting is to providing exact Impacket command syntax for a senior operator.
How It Works
- Install the agent files into your Claude Code agents directory
- Open Claude Code and describe your task naturally
- Claude automatically routes to the right specialist agent
No configuration, no commands to memorize. Just describe what you need.
Quick Demo: Advisory Mode (Tier 1)
$ claude
You: Analyze this Nmap scan and prioritize targets for our internal pentest
▸ Routing to recon-advisor agent...
## Prioritized Target Summary
### Critical Priority (Exploit Immediately)
| Host | Port | Service | Finding |
|------------|------|---------|--------------------------|
| 10.10.1.5 | 445 | SMB | SMBv1 enabled, MS17-010 |
| 10.10.1.20 | 3389 | RDP | BlueKeep (CVE-2019-0708) |
### Recommended Next Steps
1. nmap -sV --script smb-vuln* 10.10.1.5
2. crackmapexec smb 10.10.1.0/24
3. bloodhound-python -d corp.local
Quick Demo: Execution Mode (Tier 2)
$ claude
You: My authorized scope is 10.10.1.0/24. Internal network pentest.
Run a service scan on the subnet.
▸ Routing to recon-advisor agent...
Scope recorded: 10.10.1.0/24 (internal)
I'll start with host discovery, then do targeted service detection
on live hosts.
[MODERATE] Host discovery scan:
nmap -sn 10.10.1.0/24 -oN nmap_10.10.1.0_discovery_20260330_140000.txt
▸ Allow Bash command? [y/n]
Found 47 live hosts. Running service detection on the top ports...
[MODERATE] Service scan on live hosts:
nmap -sT -sV -sC --top-ports 1000 --min-rate 100 --max-rate 1000
--host-timeout 300s -iL live_hosts.txt
-oN nmap_10.10.1.0_services_20260330_140215.txt
▸ Allow Bash command? [y/n]
## Prioritized Target Summary
| Priority | Host | Port | Service | Finding |
|----------|------------|------|----------------|-----------------------|
| Critical | 10.10.1.5 | 445 | SMB 1.0 | MS17-010 likely |
| Critical | 10.10.1.20 | 3389 | RDP 10.0.17763 | BlueKeep candidate |
| High | 10.10.1.12 | 8080 | Jenkins 2.289 | Pre-auth RCE range |
Evidence saved. Want me to run SMB vulnerability scripts on 10.10.1.5 next?
Agents
Offensive Operations
| Agent | What It Does | Example Prompt |
|---|---|---|
| Engagement Planner | Plans penetration tests with phased methodology, MITRE ATT&CK technique mapping, time estimates, and rules of engagement templates | "Plan an internal network pentest for a 500-endpoint Active Directory environment with a 2-week window" |
| Recon Advisor | Parses output from Nmap, Nessus, BloodHound, and 20+ tools. Prioritizes targets, maps CVEs, and recommends specific next commands. Tier 2: can execute recon tools directly with your approval. | "Scan 10.10.1.0/24 and tell me what to hit first" |
| OSINT Collector | Open source intelligence gathering: domain recon, email harvesting, social media profiling, breach data analysis, and infrastructure mapping | "Build an OSINT profile on this target domain before our external engagement" |
| Exploit Guide | Detailed exploitation methodology covering AD attacks, web apps, cloud, and post-exploitation. Every technique includes the defensive perspective | "Walk me through AS-REP Roasting and how defenders detect it" |
| Privilege Escalation | Systematic Linux and Windows privilege escalation methodology. SUID abuse, token impersonation, service exploitation, kernel exploits, and container escape | "Here's my linpeas output, what's the fastest path to root?" |
| Cloud Security | AWS, Azure, and GCP penetration testing methodology. IAM privilege escalation, container escape, serverless exploitation, and cloud-native attack paths | "I have read-only AWS access with this IAM policy. Find privilege escalation paths" |
| API Security | REST, GraphQL, and WebSocket security testing. OWASP API Top 10, JWT attacks, OAuth exploitation, BOLA/BFLA testing, and API discovery | "Test this API for BOLA. Here's the Swagger doc and a valid JWT" |
| Mobile Pentester | Android and iOS application security testing. APK/IPA analysis, Frida hooking, SSL pinning bypass, OWASP MASTG/MASVS methodology | "Decompile this APK and check for hardcoded secrets and certificate pinning" |
| Wireless Pentester | WiFi and Bluetooth penetration testing. WPA/WPA2/WPA3 attacks, evil twin, rogue AP, enterprise wireless, and Bluetooth security | "Capture a WPA2 handshake and set up an evil twin for this corporate network" |
| Social Engineer | Phishing campaigns, pretexting, vishing, physical social engineering, and security awareness assessments for authorized red team engagements | "Design a phishing campaign for this engagement using GoPhish" |
| Vuln Scanner | Runs vulnerability scans with Nuclei, Nikto, and Nmap NSE scripts. Parses Nessus and OpenVAS results. Prioritizes by CVSS and exploit availability. Tier 2: executes scans directly with your approval. | "Run a Nuclei scan on 10.10.1.0/24 for critical and high severity CVEs" |
| Web Hunter | Web application testing with ffuf, gobuster, feroxbuster, sqlmap, and dalfox. Content discovery, parameter fuzzing, virtual host enumeration, and WAF detection. Tier 2: executes web testing tools directly. | "Fuzz directories on https://target.com and check for SQL injection on the login form" |
| Credential Tester | Password attack methodology covering Hydra, Hashcat, John the Ripper, CrackMapExec spraying, Kerbrute, and custom wordlist generation. Hash identification and cracking strategy | "I have these NTLM hashes from a SAM dump. What's the fastest cracking approach?" |
| Attack Planner | Correlates findings from all other agents into multi-step attack chains. Scores paths by probability, stealth, and business impact. Builds lateral movement maps and chain comparison matrices | "I have Nmap results, BloodHound data, and some cracked hashes. Build me the best attack chain to DA" |
| Bug Bounty Hunter | Bug bounty methodology for HackerOne, Bugcrowd, and Intigriti. Target selection, recon automation, duplicate avoidance strategies, and professional report writing that gets bounties paid | "Help me write a P1 report for this IDOR I found on HackerOne" |
| AD Attacker | Active Directory attack execution with BloodHound, Impacket, CrackMapExec, Certipy, and Kerbrute. Kerberos attacks, delegation abuse, ACL exploitation, and certificate abuse. Tier 2: executes AD tools directly with your approval. | "Kerberoast all service accounts in corp.local and crack the hashes" |
| Exploit Chainer | Autonomous exploit chaining that takes isolated |