Cool conceptually, but how exactly does the agent know when it's unsure or stuck?
> (problem description) your agent […] makes confident assumptions
> (solution description) when it’s unsure
I read this as a contradiction: in one sentence you describe the problem as an agent being confident while hallucinating and in the next phrase the solution is that the agent can ask you if it’s unsure.
You tool is interesting but you may consider rephrasing that part.
So not at all, but that doesn't mean it's not useful.
Yes, it may not need to know with perfect certainty when it's unsure or stuck, but even to meet a lower bar of usefulness, it'll need at least an approximate means of determining that its knowledge is inadequate. To purport to help with the hallucination problem requires no less.
To make the issue a bit more clear, here are some candidate components to a stuck() predicate:
- possibilities considered
- time taken
- tokens consumed/generated (vs expected? vs static limit? vs dynamic limit?)
If the unsure/stuck determination is defined via more qualitative prompting, what's the prompt? How well has it worked?
(But you could build one that does this, and ask the LLM to call it and give your MCP that data... when it feels like it)
So you'd be using this by telling the LLM to run it when it thinks it's stuck. Or needs human input.
1: I am not anything even approaching deeply knowledgeable about MCP, so please, someone correct me if I'm wrong! There do seem to be some bi-directional messaging abilities, e.g. notification, but to figure out thinking time / token use / etc you would need to have access to the infrastructure running the LLM, e.g. Cursor itself or something.
You can probably get some where by indeed running a task 1000 times and looking for outliers in the execution time or token count. But that is of minimal use and anything more advanced than that is akin to water divining.
This is little different than how I wrote this. Now it is deterministic, when I hit reply.