Epoch Protocol primary sources — 2026-05-08

Classification takeaway

Epoch Protocol is best understood as a coordination layer for competing intent solvers rather than as a simple bridge, router, or account-abstraction wallet product. The key mechanism is the SIO layer: it decomposes a user’s desired end state into sub-intents, queries multiple solver paths, weighs constraints and optimization goals, and decides how the aggregate solution should be assembled. That makes Epoch most useful as a comparison class for intent-routing systems, solver markets, account-abstraction automation, and cross-chain state-orchestration middleware.

Primary sources reviewed

1. Official site — https://epochprotocol.xyz/

   • Marketing surface describing Epoch as an intent layer / coordination layer for agentic and multichain execution.
   • Useful as first-party confirmation that the protocol sees itself as a coordination surface, not merely another bridge.

2. Docs root — https://docs.epochprotocol.xyz/

   • Best starting point for the protocol’s own framing of its architecture and intent model.
   • Useful because it ties together litepaper, overview pages, SDK docs, and contracts references.

3. Litepaper — https://docs.epochprotocol.xyz/litepaper

   • The main architectural source.
   • Defines Epoch as an intent-solver coordination layer, introduces SIOs, Hyperions, execution clients, alternate mempools, consensus, smart wallets, plugins, hooks, paymasters, and permissioned liquidity pools.

4. Litepaper Q&A: components + coordination — https://docs.epochprotocol.xyz/litepaper?ask=What%20are%20Epoch%27s%20main%20components%20and%20how%20does%20solver%20coordination%20work%3F

   • Best concise source for the protocol’s operating loop.
   • Shows the intended sequence: intent submission → SIO decomposition → Hyperion state checks → multiple solver queries → weighted assembly → consensus selection → atomic settlement.

5. Litepaper Q&A: non-custody + contracts — https://docs.epochprotocol.xyz/litepaper?ask=How%20does%20Epoch%20keep%20user%20funds%20non-custodial%20and%20what%20on-chain%20contracts%20does%20it%20use%3F

   • Important because it exposes the real settlement trust surface.
   • Says SAFE wallets keep custody, users remain the only signers, and onchain enforcement uses registry/gatekeeper contracts, plugins, hooks, paymaster support, and permissioned liquidity pools.

6. Overview page — https://docs.epochprotocol.xyz/epoch-overview.md

   • Reinforces the positioning around solver-solver coordination and observer-triggered execution.
   • Useful because it is often clearer and more concise than the full litepaper for classification work.

7. Contracts docs — https://docs.epochprotocol.xyz/use-epoch-sdk/contracts.md

   • Useful for checking how much of the architecture is conceptual versus implemented in named onchain components.
   • Helps ground the account-abstraction and settlement claims in explicit contract surfaces.

8. GitHub org — https://github.com/epochprotocol

   • Confirms an active public code surface across contracts and SDKs.
   • Worth revisiting later to map which parts of the orchestration stack are actually open versus described only in docs.

9. Epoch Intent SDK repo + README — https://github.com/epochprotocol/epoch-intents-sdk and https://raw.githubusercontent.com/epochprotocol/epoch-intents-sdk/main/README.md

   • Best public operational evidence that intents are meant to be created, signed, and submitted through a developer-facing integration surface.
   • Especially useful because it makes nonce handling, signing, and node submission concrete instead of hypothetical.

Why it stays in the active corpus

• Strong comparison class for intent-centric systems where orchestration logic, not just solver speed, determines practical power.
• Helps reason about where authority migrates when users express desired outcomes and middleware decides how to decompose, monitor, and settle them.
• More analytically interesting than a generic bridge or automation layer because the SIO layer turns routing policy itself into a control surface.