OpenClaw security assessment [pdf]
(zeroleaks.ai)56 points by nreece 4 hours ago
56 points by nreece 4 hours ago
Zeroleaks.ai is a 13 day old registration. Cool.
More interesting, looks to be from this 16yo, https://github.com/x1xhlol, https://www.lucknite.dev/
The account's stars are mostly a "system prompts" collection repo fwiw.
Trying to hustle online and writing high quality software aren't the same
Looks interesting, https://github.com/ZeroLeaks/zeroleaks
At least, I am curious about the tool
Zero mention of specific models that are being compromised makes it hard to take the numbers in this report seriously.
I do understand there's a lot of people running openclaw that don't really understand it and know what models are actually running. But we've known for a while that there are tons of older models that are pretty vulnerable, and you can hook up any model to OpenClaw, so, this data is not really that useful. Even though I totally agree that there are plenty of security risks here
Relying on the model for security is not security at all.
No amount of hardening or fine-tuning will make them immune to takeover via untrusted context
Almost all of this report is about leaking system prompts.
The OpenClaw system prompt has no measures in it at all to prevent leaking, because trying to protect your system prompt is almost entirely a waste of time and actually makes your product less useful.
As a result, I do not think this is a credible report.
Here's the system prompt right now: https://github.com/openclaw/openclaw/blob/b4e2e746b32f70f8fb...