You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
Tighten v1 security invariants and define the trusted-computer boundary #4
The v1 security documentation needs one literal, consistent trusted-computer and secret-output contract.
Required contract:
define a trusted computer operationally as a machine under the operator's exclusive control, not known or suspected to be compromised, with the OS/Python environment/codex32/bitcoin-cli/Bitcoin Core and relevant configuration trusted for the operation;
distinguish intentional secret-emitting commands from accidental disclosure so invariant 6 is literally true;
warn that shell command text can be retained even when prompted or redirected stdin keeps secrets out of argv;
keep SECURITY.md, the security model/invariants, user guidance, and CLI help consistent.
Implemented in focused PR #59. The final refreshed head ede98a8 is mergeable, has resolved review threads, passed exact-head Python-package run 643 and Bitcoin Core wallet-fixture run 30, and has a current-head Codex release-gate ACK. The refreshed wording is reconciled with #23 and the #57/#80/#81 restore contract.
Keep this issue open through human integration of #59; no automated/code-review gap remains.
Activity
docs: Define trusted computer boundary
docs: Define trusted computer boundary