Always run in sandboxes. LLMs without any backdoor tools inv...

redshift

npub1ftt05tgku25m2akgvw6v7aqy5ux5mseqcrzy05g26ml43xf74nyqsredsh

hex

ec330d081623ebcef86c859db7377cc4a1f294876fb685b9248b94c06b82c0c5

nevent

nevent1qqswcvcdpqtz867wlpkgt8dhxa7vfg0jjjrkld59hyjgh9xqdwpvp3gprpmhxue69uhhyetvv9ujuem4d36kwatvw5hx6mm9qgsy44h695tw92d4wmyx8dx0wsz2wr2dcvsvp3z86y9ddl6cnyl2ejqfwqcn9

Kind-1 (TextNote)

2026-05-03T16:10:25Z

↳ 回复 Leo Wandersleb (npub1gm7tuvr9atc6u7q3gevjfeyfyvmrlul4y67k7u7hcxztz67ceexs078rf6)

Reading messages can hardly be avoided but adding backdoor tool invocations in llm replies is even scarier but probably mitigatable.

Always run in sandboxes. LLMs without any backdoor tools invocations always are notorious in screwing things up like leaking private keys/deleting data.

And we are working towards mitigating man in the middle attacks. It's an interesting problem.

原始 JSON

{
  "kind": 1,
  "id": "ec330d081623ebcef86c859db7377cc4a1f294876fb685b9248b94c06b82c0c5",
  "pubkey": "4ad6fa2d16e2a9b576c863b4cf7404a70d4dc320c0c447d10ad6ff58993eacc8",
  "created_at": 1777824625,
  "tags": [
    [
      "e",
      "b65a21d7d19bcb960876ad5cd6649da827f09f4056fba0e1ca2d0cc77907a4fd",
      "wss://nostr.wine/",
      "root",
      "f985d309197c805e1719c73185b574fc3ee407d7c1b6157dee99c6ace2599bbb"
    ],
    [
      "e",
      "faca766c836e4290a8b79fcff325917949988946f21c43907ee9d431045f0f85",
      "wss://relay.primal.net/",
      "reply",
      "46fcbe3065eaf1ae7811465924e48923363ff3f526bd6f73d7c184b16bd8ce4d"
    ],
    [
      "p",
      "8bf629b3d519a0f8a8390137a445c0eb2f5f2b4a8ed71151de898051e8006f13"
    ],
    [
      "p",
      "2efaa715bbb46dd5be6b7da8d7700266d11674b913b8178addb5c2e63d987331"
    ],
    [
      "p",
      "f985d309197c805e1719c73185b574fc3ee407d7c1b6157dee99c6ace2599bbb"
    ],
    [
      "p",
      "d22bcdb1d2505bc060f65a86cd5f20532ec5f3e41c06aea9edac39f6ba76fd6e"
    ],
    [
      "p",
      "46fcbe3065eaf1ae7811465924e48923363ff3f526bd6f73d7c184b16bd8ce4d"
    ]
  ],
  "content": "Always run in sandboxes. LLMs without any backdoor tools invocations always are notorious in screwing things up like leaking private keys/deleting data. \n\nAnd we are working towards mitigating man in the middle attacks. It's an interesting problem.",
  "sig": "d6ad5c1e5293d47f7bf51ab279b31b5616af332bef4abde9f866b27e8d3a4b61137eecf40da8830a4a6d313dd9fe9e13538f43d7f19d2ad3be2b05abc8d3b88b"
}