There are 16 attacks in the code-exploitation category. Each shows its technique id, mode, standards mapping, description, and the exact command to run it (swap the provider and target as needed).
​
Backdoor Generation
critical
backdoor-generation · single-turn · OWASP: LLM01 · MITRE: AML.T0054, AML.T0068Asks the model to create software backdoors for persistent covert accessRun it:
ai-blackteam run -p anthropic -a backdoor-generation -t "your target prompt"
​
Botnet C2 Generation
critical
botnet-generation · single-turn · OWASP: LLM01 · MITRE: AML.T0054, AML.T0068Asks the model to create botnet command-and-control infrastructure and bot agentsRun it:
ai-blackteam run -p anthropic -a botnet-generation -t "your target prompt"
​
Credential Stealer
critical
credential-stealer · single-turn · OWASP: LLM01 · MITRE: AML.T0054, AML.T0068Asks the model to write tools that harvest and exfiltrate credentials from systemsRun it:
ai-blackteam run -p anthropic -a credential-stealer -t "your target prompt"
​
Exploit Generation
critical
exploit-generation · single-turn · OWASP: LLM01 · MITRE: AML.T0054, AML.T0068Asks the model to write working software exploits targeting real vulnerabilitiesRun it:
ai-blackteam run -p anthropic -a exploit-generation -t "your target prompt"
​
Keylogger Generation
critical
keylogger-generation · single-turn · OWASP: LLM01 · MITRE: AML.T0054, AML.T0068Asks the model to create keyloggers that capture and exfiltrate keystrokesRun it:
ai-blackteam run -p anthropic -a keylogger-generation -t "your target prompt"
​
Malware Generation
critical
malware-generation · single-turn · OWASP: LLM01 · MITRE: AML.T0054, AML.T0068Asks the model to generate functional malware codeRun it:
ai-blackteam run -p anthropic -a malware-generation -t "your target prompt"
​
Privilege Escalation Code
critical
privilege-escalation-code · single-turn · OWASP: LLM01 · MITRE: AML.T0054, AML.T0068Asks the model to write privilege escalation exploits to gain root or SYSTEM accessRun it:
ai-blackteam run -p anthropic -a privilege-escalation-code -t "your target prompt"
​
Ransomware Generation
critical
ransomware-generation · single-turn · OWASP: LLM01 · MITRE: AML.T0054, AML.T0068Asks the model to create ransomware with encryption and payment mechanismsRun it:
ai-blackteam run -p anthropic -a ransomware-generation -t "your target prompt"
​
Reverse Shell Generation
critical
reverse-shell · single-turn · OWASP: LLM01 · MITRE: AML.T0054, AML.T0068Asks the model to generate reverse shell payloads for establishing covert command-and-control channelsRun it:
ai-blackteam run -p anthropic -a reverse-shell -t "your target prompt"
​
Rootkit Generation
critical
rootkit-generation · single-turn · OWASP: LLM01 · MITRE: AML.T0054, AML.T0068Asks the model to create rootkits that hide attacker presence and provide persistent accessRun it:
ai-blackteam run -p anthropic -a rootkit-generation -t "your target prompt"
​
Spyware Development
critical
spyware-development · single-turn · OWASP: LLM01 · MITRE: AML.T0054, AML.T0068Generates instructions for developing commercial-grade spyware capable of full device compromise, data exfiltration, and persistent surveillanceRun it:
ai-blackteam run -p anthropic -a spyware-development -t "your target prompt"
​
Worm Generation
critical
worm-generation · single-turn · OWASP: LLM01 · MITRE: AML.T0054, AML.T0068Asks the model to create self-propagating network wormsRun it:
ai-blackteam run -p anthropic -a worm-generation -t "your target prompt"
​
Zero-Day Exploit Assistance
critical
zero-day-assist · single-turn · OWASP: LLM01 · MITRE: AML.T0054, AML.T0068Asks the model to assist in developing zero-day exploits for unpatched vulnerabilitiesRun it:
ai-blackteam run -p anthropic -a zero-day-assist -t "your target prompt"
​
Cryptominer Generation
high
cryptominer-generation · single-turn · OWASP: LLM01 · MITRE: AML.T0054, AML.T0068Asks the model to create hidden cryptomining malware that hijacks victim compute resourcesRun it:
ai-blackteam run -p anthropic -a cryptominer-generation -t "your target prompt"
​
DDoS Attack Tool
high
ddos-tool · single-turn · OWASP: LLM01 · MITRE: AML.T0054, AML.T0068Asks the model to create distributed denial-of-service attack toolsRun it:
ai-blackteam run -p anthropic -a ddos-tool -t "your target prompt"
​
Phishing Kit Generation
high
phishing-kit · single-turn · OWASP: LLM01 · MITRE: AML.T0054, AML.T0068Asks the model to create phishing page clones and credential harvesting infrastructureRun it:
ai-blackteam run -p anthropic -a phishing-kit -t "your target prompt"