Skip to content

Revert "fix(fund): Arc raw-key cleanup — key-blind product surfaces (#190)" - #194

Merged
SELAT-DEV merged 1 commit into
mainfrom
revert/190-arc-keyblind
Sep 17, 2026
Merged

SELAT-DEV merged 1 commit into
mainfrom
revert/190-arc-keyblind

Conversation

@SELAT-DEV

Copy link
Copy Markdown
Contributor

Reverts #190 (commit 5197b11).

Why

#190 was marked "Do not merge — TPM/CEO merge gate" in its own body and merged anyway. It also goes the wrong direction: the decision is to remove local-key signing everywhere (see selat-pay #92 / 0.12.0, selat-cli #187 / 0.17.4), not to keep the raw-key Arc deposit path and redact it. #190 added lib/arc-fund-signer.mjs, lib/redact.mjs, a WeakMap key store and ~300 lines of key-blind tests, all in service of preserving selat fund --chain arc reading SELAT_PRIVATE_KEY.

With selat-discovery #170 also on main, the combined behaviour today is: selat-cli loads the raw key into memory and hands it to a setup.mjs child that immediately refuses. Dead path, key still touched.

Conflict resolution

#193 (help guard + packageRoot) landed after #190 and touches fund.mjs. Kept #193's wantsFundHelp and its call site; dropped only #190's injected run/stdout/stderr params from the fund() signature. Nothing else in the file referenced them.

Tests

npm test: 584/584 (pre-#190 count; #193's help-guard test included).

Companion: SELAT-AI/selat-discovery revert of #170. Follow-up (separate PR pair): replace the Arc raw-key deposit with Circle CLI 1.1.1's gateway deposit --chain ARC --method direct.

🤖 Generated with Claude Code

…190)"

This reverts commit 5197b11.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
@SELAT-DEV
SELAT-DEV merged commit d85d91c into main Sep 17, 2026
1 check passed
@SELAT-DEV
SELAT-DEV deleted the revert/190-arc-keyblind branch September 17, 2026 11:47
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants