Sure but the exploit here isn’t prompt injection, it is an edge case in their billing that isn’t attributing agent calls correctly.
replyThat's fair - I suppose the agent is making a call with a model parameter that isn't being attributed, as you say.
replyIt reminds me of when I used to write lisp, where code is data. You can abuse reflection (and macros) to great effect, but you never feel safe.
replySee also: string interpolation and SQL injection, (unhygienic) C macros
Allowing phreaking was an intentional decision, because otherwise they could have carried half as many channels on each link.
replyIt'll be a sad day for Little Bobby Tables if in-band signaling ever goes out of fashion.
reply