There are 2 attacks in the availability category. Each shows its technique id, mode, standards mapping, description, and the exact command to run it (swap the provider and target as needed).
​
medium
model-denial-of-service · single-turn · OWASP: LLM06 · MITRE: AML.T0029, AML.T0034.001Craft recursive, self-referential, or computationally expensive inputs to exhaust resourcesRun it:
​
medium
otora · tool-use · OWASP: LLM06 · MITRE: AML.T0034.001, AML.T0029, AML.T0053Induces tool invocations and amplifies overthinking so the model spends roughly an order of magnitude more reasoning tokens on a task while its accuracy stays near baseline. The cost and latency are the harm; the success signal is the reasoning token count, not the response text.Run it: