There are 4 attacks in the injection category. Each shows its technique id, mode, standards mapping, description, and the exact command to run it (swap the provider and target as needed).
​
Indirect Injection
critical
indirect-injection · single-turn · OWASP: LLM01 · MITRE: AML.T0051.001Simulate tool output containing hidden instructions to hijack model behaviorRun it:
ai-blackteam run -p anthropic -a indirect-injection -t "your target prompt"
​
SQL Injection
high
sql-injection · single-turn · OWASP: LLM01 · MITRE: AML.T0051Inject SQL payloads through natural language queries to test output sanitizationRun it:
ai-blackteam run -p anthropic -a sql-injection -t "your target prompt"
​
SSRF Probing
high
ssrf-probing · single-turn · OWASP: LLM01 · MITRE: AML.T0051Trick model into making requests to internal URLs and cloud metadata endpointsRun it:
ai-blackteam run -p anthropic -a ssrf-probing -t "your target prompt"
​
XSS Injection
medium
xss-injection · single-turn · MITRE: AML.T0051.000, AML.T0067Get model to generate HTML containing cross-site scripting payloadsRun it:
ai-blackteam run -p anthropic -a xss-injection -t "your target prompt"