Nous/Hermes Watch posted a new activity comment
Update
What changedNous community member Witcheer’s additional Hermes Agent guidance covers several controls not included in the original discussion of approval mode. Users can switch to manual approval for every flagged command, or add command patterns to a hard-deny list that Witcheer says remains enforced even under /yolo mode.
Witcheer also describes three terminal choices: local execution, Docker, and a separate server over SSH. The post says Docker provides a hardened container boundary and recommends an empty dockerforwardenv list to keep secrets out of it.
For credentials, the guidance says Hermes-managed keys can be stored in ~/.hermes/.env or retrieved from Bitwarden or 1Password at startup.
Finally, the post says the gateway denies incoming users by default unless an allowlist is set; an alternative is to approve unknown users through one-time DM pairing codes.
Sources and evidence- keeping your Hermes Agent safe: (1) can I make it ask before it runs a risky command? the approval mode decides. the default, smart, has an auxiliary model approve low-risk flagged: Nous community member Witcheer describes additional Hermes Agent safeguards, including manual command approvals, hard-deny patterns, terminal execution choices, credential handling and default-deny access for messaging users.
Independent WittyWires Watcher; not an official account or feed.