Wrong. This is largely a solved problem with “auto review” permissions mode, which uses another LLM to review all tool calls that could potentially be dangerous. That solution isn’t perfect of course but it is reliable and usable at the point. This is available in Claude code, codex, copilot, and a bunch of others.
I find auto review mode mandatory and refuse to consider any harness that doesn’t support it.
Opencode doesn’t support it. And there are no good plugins to do it either, the most popular had only six stars last I checked.
Auto review often refuses to do things I directly ask the LLm to do, the other day it refused to go to a personal blog and copy the contents for a prototype so I could test a new design. Have had it refuse to use test data on numerous occasions and only tell me when I manually checked and realized it wouldn’t do it
Yes, that was Google’s justification for killing MV2. It isn’t wrong, of course, but I insist on defining my own security posture and I prioritize adblocking.
At least in the self-hosted LLM inference engines, you have to pre-allocate space for the maximum amount of context you want to allow for each parallel session. By using a lower maximum, you don't have to allocate as much VRAM for each session, allowing more usage for the same amount of hardware. Thus, cheaper.
This was done because they were charging the 100%/50% premium for long-context over 272k to Codex users, even though they set 372k as the default. It cost people a LOT of money.
I find auto review mode mandatory and refuse to consider any harness that doesn’t support it.
Opencode doesn’t support it. And there are no good plugins to do it either, the most popular had only six stars last I checked.
reply