Official docs overview describing Cork as a programmable risk layer for onchain assets, with the core Collateral Asset / Reference Asset pairing and cPT/cST token split.
Public GitHub mirror of the docs content; useful because it exposes the same overview text and use-case framing even when some docs pages are behind browser protections.
Best single technical source surfaced in this pass for architecture and control surfaces. Describes the singleton CorkPoolManager, DefaultCorkController, WhitelistManager, ConstraintRateAdapter, CorkAdapter, and the lifecycle operations around deposit/mint, exercise/swap, unwind, and post-expiry redemption.
Earlier core-contract repository showing the protocol lineage around depeg swaps and formal-verification/test workflow.
Key takeaways from primary materials
Cork is not just “depeg insurance”; it is a reusable redemption-liquidity primitive where depositing collateral mints two separable claims: principal (cPT) and swap/coverage (cST).
The important reusable analogy is a tokenized redemption right or swap-like liquidity backstop, closer to a programmable CDS / protected-exit rail than to a conventional insurance mutual.
The Phoenix architecture makes governance and control surfaces legible: DefaultCorkController authorizes pools, controls fees and pauses, and manages whitelist access, while WhitelistManager can enforce both global and market-specific access.
The presence of ConstraintRateAdapter, oracle wrapping, and a protected CorkAdapter layer shows the protocol is explicitly trying to bound rate/oracle risk and user execution risk, not only trade risk claims.
The strongest comparison classes in the corpus are Cork’s own redemption-liquidity niche, Risk Protocol’s tokenized-risk framing, and insurance-adjacent systems that isolate one awkward slice of onchain risk into a separately tradeable claim.