Is this agent safe to trust?
Scan any AI agent on 0G. Sentinel checks it against real sanctions and scam-blocklist intelligence, reads its contract code, and seals the verdict on-chain so anyone can check your work.
Every scan screens OFAC sanctions, Chainalysis, and ScamSniffer, and disassembles the agent's contract bytecode, then seals the full evidence on 0G Storage. A verdict you can check, not a vibe you have to trust.
Check the work yourself.
Every verdict is sealed on 0G with its full evidence. Browse recent scans, or re-open any past proof and reproduce it yourself — no login, no paid inference.
Honest about its limits.
Sentinel scores how risky an agent looks right now — it is not a guarantee of safety. It ties the verdict to the agent's real on-chain identity and wallet, and both the automated and AI checks can still miss things.
Every verdict is versioned and kept, so history can't be quietly rewritten. Older V1 records stay clearly labeled and never count as a current pass.
Verify a past proof yourself →