This message was deleted.
# openhands-cloud
d
This message was deleted.
🙌 1
p
What would stop folks from abusing the system and use all your llm tokens to solve their own problems? I wonder how LangChain safeguards https://chat.langchain.com/ from this(?) What if OpenHands, in the background (you'd be controlling it), would regularly maintain the "OpenDeepWiki"?
l
It would be using their tokens.
They’d need to log in to be able to use it
(although I don’t expect that answering questions takes many tokens)
p
can't folks already do that via
llms.txt
MCP approach? https://openhands-ai.slack.com/archives/C08CR9XRQ3W/p1746139315232899?thread_ts=1746060945.051369&cid=C08CR9XRQ3W (use their tokens to query any project)
l
Yeah, this is just a convenience that would allow them to quickly ask a question about the codebase instead of reading the docs
p
I did use https://help.openai.com/en/articles/11145903-connecting-github-to-chatgpt-deep-research against OpenHands, it's quite good I like the idea of using OpenHands to understand OpenHands 😄
l
Gotta dogfood everything 🙂
And OpenHands is quite good. I haven’t tried deep research against the repo, but OpenHands is at least as good as DeepWiki, usually better (but sometimes slower)
🔥 1
w
Allowing links to kick of an openhands conversation could potentially be an attack vector. Definitely need to be careful
l
Yeah, I agree, I don’t think we should allow starting OpenHands work. But I think opening a conversation on a particular repo might be OK?
w
probably...it could maybe have a toxic microagent
maybe we need a "do you trust this" similar to claude code
l
yeah sounds good