January 30, 2026
ClickFix → ClickBroke
ClawdBot Skills ganked all my crypto
Fake ClawdBot crypto “skills” snatch wallets, commenters go savage
TLDR: Fourteen fake ClawdBot add-ons posed as crypto helpers and installed info-stealing malware, snatching keys and passwords. Comments roast careless installs and hype-driven AI, warning this is a textbook supply-chain risk and a wake-up call for anyone giving chatbots access to sensitive accounts.
ClawdBot, the buzzy local AI assistant, just face-planted into a full-on crypto heist drama. Researchers say 14 flashy “skills” were uploaded to ClawHub and GitHub in late January, posing as trading tools for ByBit, Polymarket, and more. Instead of profits, they pushed sketchy zip downloads like “PolymarketAuthTool.zip” and sneaky commands that hoover up wallet keys, passwords, and exchange logins—macOS and Windows both got hit. All roads led back to the same control server, meaning this wasn’t a one-off; it was organized.
The comments lit up like a Christmas tree. The harsh crowd went full gladiator: _se snapped, “Anyone dumb enough to run this on their computer deserves it,” while lpcvoid called it peak “self‑pwn.” The policy nerds chimed in with this_user warning it’s a classic supply‑chain mess—installing add‑ons from random registries is basically inviting trouble. And the conspiracy‑flavored takes? dispersed side‑eyed the hype machine, calling it “pushed so hard to normies” and “100% predictable.”
There’s some “told you so” energy too, with nods to earlier ClawdBot security warnings and folks wondering why the official registry didn’t scan listings when payloads were visible in plain text. The meme of the moment: ClickFix → ClickBroke. The vibe: clown car meets crypto wallet. The lesson: don’t hand your keys to a chatbot dressed like a day trader.
Key Points
- •14 ClawdBot skills were uploaded Jan 27–29, 2026, to ClawHub and GitHub, with 12 confirmed as malicious.
- •The skills impersonate crypto-trading tools and target users of Claude Code and Moltbot on macOS and Windows.
- •All malicious skills share a C2 IP address (91.92.242.30) and use social engineering to execute payloads.
- •Exfiltrated data includes exchange API keys, wallet private keys, SSH credentials, and browser passwords.
- •Investigators found no evidence of security scanning on ClawHub; some payloads were visible in SKILL.md files.