Skip to content
Orryn
Dependency repair · testnet prototype

A published fix is only the start.Prove each deployment is repaired.

Orryn helps infrastructure suppliers and protocol teams coordinate known dependency repairs across customer-owned smart-contract deployments—with clear checks, owner-controlled changes, and evidence tied to the state that was observed.

release policy → deployment evidenceno customer key custody

Current prototype supports one documented oracle-adapter and proxy pattern; unsupported paths are reported as unsupported, not silently accepted.

Made for the handoff between

Supplier security teamsCustomer protocol teamsAuthorized owners

The gap

One supplier release. Many customer-owned deployments. Who checked that they all changed?

A release announcement says a repair exists. It does not establish which customer deployments use it, whether their configuration is safe, or whether the agreed behavior still works.

01

A release note is not deployment coverage

Publishing a repaired component does not update the customer-owned contracts that depend on it.

02

A version string is not the whole story

A deployment can use the expected code and still have a wrong route, unsafe setting, or unmet behavior check.

03

The customer controls the repair

Suppliers can explain and prepare a change, but only the customer’s owner, multisig, or governance can authorize it.

04

Evidence and funding are disconnected

When a supplier funds remediation, it needs a clear way to tie each eligible deployment’s result to its grant rules.

Repair workflow

A repair process with evidence at each handoff.

Orryn coordinates and explains the work. The customer remains in control of its deployment and its keys.

01Supplier

Publish the repair policy

The supplier defines the affected deployment scope and the conditions an acceptable repair must meet.

02Read only

Inspect the customer’s deployment

Orryn reads the supported onchain dependency path and records the chain state it observed.

03Evidence

Explain what does not match

Each required check reports pass, fail, unknown, or unsupported, with its expected and observed evidence.

04Customer controlled

Customer authorizes its own repair

Orryn can prepare and simulate a proposed transaction. The customer’s authorized wallet decides whether to execute it.

05Arc · planned end-to-end

Recheck, then apply campaign rules

A fresh result can be tied to a funded Arc campaign. The contract—not a page badge—controls grant settlement.

Architecture

Observe on the customer’s chain. Settle the campaign on Arc.

The website and verifier run as regular web services. The deployment being inspected stays on its source chain; Arc is the destination for campaign USDC rules. Chainlink CRE is intended to coordinate evidence and chain facts, and is still an integration in progress.

Source chain

Customer deployment

Proxy, dependency route, configuration, and current block state.

Orryn verifier

Checks and evidence

Runs the named supported checks and records a scoped outcome.

Arc settlement

Campaign contract

Applies funded USDC eligibility and claim rules onchain.

Chainlink CRE: planned evidence/fact orchestration.End-to-end workflow not live yet

Evidence, with boundaries

Every result should answer: what was checked, and against what?

01

Observe before you conclude

Results are bound to a named deployment and observed chain snapshot. A missing or stale observation is not a pass.

02

Code and configuration both matter

The supported check looks at the dependency route, implementation, required configuration, and agreed behavior tests.

03

The owner stays in control

Orryn does not hold customer upgrade keys or silently change a customer’s contract.

04

Evidence has a boundary

A verifier’s signed report identifies who attested to the result; it is not a cryptographic proof that arbitrary code is secure.

Orryn is not a universal scanner or an automatic upgrader. The current prototype covers one documented oracle-adapter/proxy pattern; other patterns need explicit support or manual review.

Repair with accountability

Know which deployments are repaired—and what still needs attention.

Start with the repair policy, the customer-owned deployment, and the evidence needed to call the work done.