LLM10: Unbounded Consumption
Resource exhaustion via token bombs, expensive tool calls, or runaway agent loops driving cost or DoS.
Last reviewed October 2026
Why LLM10 evidence is hard
Cost and availability are the same failure seen from two budgets. Ten probes drive it - token bombs, latency amplification, unbounded output, distillation-shaped extraction, runaway fan-out - scored against per-probe ceilings on output tokens, latency and spend. The throughput harness is gated separately: owner or admin only, capped at one run per endpoint per hour, and refused with an audit entry before anything is queued, because a test that exists to exhaust a target needs its own brake.
How Penaxtra maps to LLM10
The LLM10 identifier is written onto the finding at the moment it is created, not mapped in afterwards, so the JSON and PDF exports come out already in your assessor's structure. Every finding carries the probe that was sent, a SHA-256 digest of the response it provoked, the tier that decided the verdict and its status history from first observation to close. Identity survives a re-scan: a control you closed and that comes back reopens under the same row with a reopen count, rather than arriving as a new issue.
LLM10 capabilities
Findings tagged with the OWASP LLM Top 10 LLM10 identifier
Penaxtra severity for this control (medium)
Cross-framework identifiers attached to the same finding where controls overlap
PDF and JSON evidence export with the control identifier attached
LLM10 control coverage
Findings for LLM10 carry the OWASP LLM Top 10 LLM10 identifier along with the corresponding control identifiers in the other frameworks Penaxtra maps, so one result is reflected across each mapped framework.
Frequently asked
What is LLM10 (Unbounded Consumption)?
Resource exhaustion via token bombs, expensive tool calls, or runaway agent loops driving cost or DoS. It is an OWASP LLM Top 10 control; Penaxtra assesses it at medium severity.
How does Penaxtra test for LLM10?
The LLM10 identifier is written onto the finding at the moment it is created, not mapped in afterwards, so the JSON and PDF exports come out already in your assessor's structure. Every finding carries the probe that was sent, a SHA-256 digest of the response it provoked, the tier that decided the verdict and its status history from first observation to close. Identity survives a re-scan: a control you closed and that comes back reopens under the same row with a reopen count, rather than arriving as a new issue.
Does a finding for LLM10 help with an audit?
Each finding is tagged with the OWASP LLM Top 10 LLM10 identifier and exported in the PDF and JSON evidence pack, so it appears on the auditor control list with the identifier already attached.