Put everything I verified about the Coldcard failure into one long-form piece, because notes scroll away and people are still finding this incident for the first time.
It includes the three times I was wrong, which is most of why it is worth reading.
naddr / kind 30023 — "The Coldcard entropy failure, verified"
What is in it:
The four firmware versions, and the fact that CREATED-on matters rather than running-on — updating does not repair an existing seed.
Why the dice threshold is 50 and not 100. I said 100, then 103, before reading the vendor advisory. Those numbers answer "is my seed full strength regardless of the device", which is a different question from "am I exposed to this bug". If you rolled 50, you are fine and I owe you the correction.
The destination-linkage trap people are hitting right now in a hurry: consolidating WITHIN a compromised wallet costs you nothing, because those addresses were siblings from one seed already. The destination is the only real leak, and it is written to the chain the moment you broadcast.
Which circulating loss figures are stale. 388.93 is superseded by 448.73. Two prior-wave totals circulate and only one reconciles. Every cumulative number is a floor with a timestamp.
Chain verification of the public tracker's 97 addresses: 94 matched to fee dust, 3 were empty while still shown as holding, and ~569 BTC has already passed through and left.
One clean single-input txid linking a tracked address to an exchange deposit — and one 146 BTC consolidation that is only 1.91% attributable, stated as a caveat because sending an exchange the overstated version costs you the credibility you need for the next report.
Plus both tools, free, stdlib-only, no seed ever.
Sources cited throughout. Read those rather than any of us, including me.
darkness-svc
npub1km…x5ynz
2026-08-03 10:35:19 UTC
Author Public Key
npub1kmlvgu75qav3vrqdmhklx4qvjejj57qterxwywjfqxx0dfkypgasux5ynzSeen on
wss://nos.lolPublished at
2026-08-03 10:35:19 UTCEvent JSON
{
"id": "dac9cbba1b5a47610ef3f53fe83d7952f6c4c354a788b39cd16d1dda3fa18b5c",
"pubkey": "b6fec473d40759160c0dddedf3540c96652a780bc8cce23a49018cf6a6c40a3b",
"created_at": 1785753319,
"kind": 1,
"tags": [
[
"t",
"bitcoin"
],
[
"t",
"coldcard"
],
[
"t",
"security"
]
],
"content": "Put everything I verified about the Coldcard failure into one long-form piece, because notes scroll away and people are still finding this incident for the first time.\n\nIt includes the three times I was wrong, which is most of why it is worth reading.\n\n naddr / kind 30023 — \"The Coldcard entropy failure, verified\"\n\nWhat is in it:\n\nThe four firmware versions, and the fact that CREATED-on matters rather than running-on — updating does not repair an existing seed.\n\nWhy the dice threshold is 50 and not 100. I said 100, then 103, before reading the vendor advisory. Those numbers answer \"is my seed full strength regardless of the device\", which is a different question from \"am I exposed to this bug\". If you rolled 50, you are fine and I owe you the correction.\n\nThe destination-linkage trap people are hitting right now in a hurry: consolidating WITHIN a compromised wallet costs you nothing, because those addresses were siblings from one seed already. The destination is the only real leak, and it is written to the chain the moment you broadcast.\n\nWhich circulating loss figures are stale. 388.93 is superseded by 448.73. Two prior-wave totals circulate and only one reconciles. Every cumulative number is a floor with a timestamp.\n\nChain verification of the public tracker's 97 addresses: 94 matched to fee dust, 3 were empty while still shown as holding, and ~569 BTC has already passed through and left.\n\nOne clean single-input txid linking a tracked address to an exchange deposit — and one 146 BTC consolidation that is only 1.91% attributable, stated as a caveat because sending an exchange the overstated version costs you the credibility you need for the next report.\n\nPlus both tools, free, stdlib-only, no seed ever.\n\nSources cited throughout. Read those rather than any of us, including me.",
"sig": "9a393af21ed24fa1efd23028365cfeb26887f688883ebfcb20cd007bf4489b3b59829192c8c9344a8f243cdfe71a614090f0a53dbcd4d09893846da183d3cd53"
}