Skip to content

Rework the figures and the D2 seizure scope in the cross-chain stablecoin RI - #102

Merged
0xNeshi merged 5 commits into
mainfrom
ccs-architecture-fixes
Aug 28, 2026
Merged

Rework the figures and the D2 seizure scope in the cross-chain stablecoin RI#102
0xNeshi merged 5 commits into
mainfrom
ccs-architecture-fixes

Conversation

@0xNeshi

@0xNeshi 0xNeshi commented Aug 28, 2026

Copy link
Copy Markdown
Collaborator

Figures

One 45-line sequence diagram at the top of section 3 covered inbound bridging,
the refund, and outbound redemption, so no flow's prose had a figure beside it.
Split into three, each next to the text it describes, plus a new one for the D2
seizure path. Two outbound arrows were wrong: the redemption attestation is a
Canton contract the attester set observes, and the attesters sign it off-ledger
rather than submitting to the escrow themselves.

Section 2 now states the rail's organizing idea before its three tables, and the
15-row role table is split into Canton parties and off-ledger actors.

D2 seizure scope

The part worth reviewing. Section 3.6 claimed the seizure sweep reaches a
settled holding; 1.3 listed that sweep as remaining work.

Resolved toward what the token standard defines: the standard gives the owner's
side of every movement to the owner and leaves the provider-versus-owner
authorization split to each registry, so a forced transfer over a settled holding
is a choice a registry defines for itself. This rail asks for none, which keeps
it usable by any CIP-0112 registry.

So D2 acts on an allocation and the holdings it locks, and the rail offers no
on-ledger control over wTOK once it settles. D3 no longer calls credited value
"D2's surface", and 4.3 leads with the bound this earns: neither the admin key
nor the Custodian key reaches a credited holding.

0xNeshi and others added 5 commits August 28, 2026 10:22
The cross-chain stablecoin architecture carried one 45-line sequence
diagram for inbound bridging, the inbound refund, and outbound
redemption. It sat at the top of section 3, so no flow's prose had a
figure beside it.

Split it into three, each placed with the text it describes:

- inbound credit, before the numbered steps in section 3.1;
- inbound refund, after the refund-exclusivity rules in section 3.1,
  now showing the expiry condition, the quorum's read of the
  credited-lock registry, and the escrow's reuse of the redemption
  verifier;
- outbound redemption, before the numbered steps in section 3.3, with
  the burn and the claim creation in one shaded transaction.

Correct two outbound arrows. The redemption attestation is a Canton
contract that the attester set observes, so the gateway creates it
rather than sending it to the attesters. The attesters sign it
off-ledger with their external-chain keys and hand it to a submitter;
they never submit to the escrow themselves.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Section 2 opened with two facts and then three dense tables, so a reader
met fifteen roles before any statement of what holds them together.

State the spine instead: an attester signature and a lock's nonce are the
only things that cross the chain boundary, the settlement batch is the
hub, and supply enters at the attested mint and leaves at the burn. Then
say what each of the three subsections contributes.

Both facts from the old opener are kept. The single Token Standard V2
registry moves into the hub sentence, and the compliance attestation and
the identity credential are now named with the action each one gates.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
The fifteen roles interleaved Canton parties, off-Canton processes, and
an external-chain contract, so a reader building a party model had to
filter the table row by row.

Group them into two tables: the ten Canton parties, then the four
off-ledger processes and the lock escrow. Each row's text and the
relative order within each group are unchanged.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
The seizure scope rested on a contradiction. Section 3.6 claimed the same
sweep reaches a holding that already settled, while the component status
listed that sweep as remaining work.

Resolve it toward what the token standard defines. The standard gives the
owner's side of every asset movement to the owner and leaves the
provider-versus-owner authorization split to each registry, so a forced
transfer over a settled holding is a choice a registry defines for
itself. The rail asks for no such choice, which keeps it usable by any
CIP-0112 registry.

State the boundary in section 3.6, record it in the scope table, and
correct the three statements that depended on the wider reading: D3 no
longer calls credited value D2's surface, the redemption path names each
path's authority instead of saying the capability "only unlocks", and the
compromised-key row now leads with the fact that neither key reaches a
credited holding.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
@0xNeshi 0xNeshi self-assigned this Aug 28, 2026
@0xNeshi 0xNeshi changed the title Ccs architecture fixes Rework the figures and the D2 seizure scope in the cross-chain stablecoin RI Aug 28, 2026

@ericnordelo ericnordelo left a comment

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM!

@0xNeshi
0xNeshi merged commit 1de349f into main Aug 28, 2026
6 checks passed
@0xNeshi
0xNeshi deleted the ccs-architecture-fixes branch August 28, 2026 09:51
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.

3 participants