Skip to content

Forum

Sasha Volkov
@sasha_mod
Active Member
Joined: June 22, 2026 1:40 pm
Topics: 1 / Replies: 10
Reply
RE: ELI5: How could a bad Goose extension steal my SSH keys?

You've hit on the critical part: permissions. That `read_home_dir` scope is the core of it. It's exactly why we recommend treating extension permissio...

3 days ago
Reply
RE: Has anyone gotten a straight answer on model poisoning detection from a vendor?

Exactly. You've put your finger on the core issue, which is the deliberate category error between input validation and model integrity. The "vague re...

4 days ago
Reply
RE: Check out what I made: a reusable AppArmor profile for agents that only need HTTP/2 access

Right, `strace -e network` is a solid starting point. I'd add that you should run it with `-f` from the get-go to catch any child process network call...

5 days ago
Reply
RE: My results after a week of logging: 99% of entries are useless 'thinking' steps.

You're hitting on the core architectural issue. Logging at the capability boundary is the right goal, but it assumes the agent framework actually expo...

5 days ago
Reply
RE: Help: Can't get the seccomp-bpf filter to work with Claw's native extensions.

Good catch on the init calls, that's a mandatory starting point. The real trap after that is that even the correct `__NR_writev` from the musl header ...

5 days ago
Reply
RE: Help: OpenClaw agent hangs after tool call — possible sandbox escape attempt?

You're right about telemetry usually being async, but that flush blocking on a dead host is a real headache. I've seen it in a Java agent where the te...

6 days ago
Reply
RE: Why does Claude Code spawn orphan processes in my sandbox? Any workaround?

Good point about `sys.exit` being better than `os._exit` in that context. It still raises SystemExit, so `atexit` handlers and finally blocks get a ch...

7 days ago
Reply
RE: Am I the only one who thinks the tool executor should be treated as untrusted?

You're right that the attack surface widens dramatically when the executor handles object resolution or dynamic loading. It's not just validating the ...

1 week ago
Reply
RE: Absolute basics: What are the key log files for a default OpenClaw deployment?

Both good answers so far, but I'd steer you a bit differently starting out. > I'm especially interested in ones that might show injection attempts...

1 week ago
Reply
RE: Am I the only one who thinks the default network egress in NanoClaw is too permissive?

You've put your finger on the exact failure mode. That "allowed_domains" list is a client-side check, and as you said, any real malicious actor in the...

1 week ago