Your Crypto Coterminal
Search, Understand, and Act. Together.
It’s time to smarten up crypto agents and humans by dumbing down smart contracts.
We’re building a collaborative terminal (coterminal) for contracts that allows you to work with agents to compose existing onchain contracts into offchain and onchain outputs. Need to build a report that goes deeper than displaying the number of transactions and users? Create your own devrel agents to help builders develop on top of your protocol? Want to have an AI agent monitor and research the most popular Uniswap v4 hooks deployed? Tired of waiting for that defi agent or app to integrate that sequence of pool actions for you? Herd is making this all possible.
Exploring onchain has long been a handicapped game for users in a massively multiplayer world, with little visibility into what actually happens or is possible with contracts onchain. Herd enables you and your team (human or AI) to search, understand, and act together. Think of our coterminal as a continuous evolution of block explorers, one that also connects powerful queries from data platforms like Dune and ecosystem discourse from social platforms like Farcaster.
We’ve raised $1.8m from Semantic, Archetype, and First Commit to build Herd. Keep reading to learn why.
The Value Hidden Under the App Layer
For years, the crypto community has chattered about the “app layer.” You have probably seen some variation of the application stack made up of a blockchain, smart contracts, node RPCs, and an interactive frontend app. Teams build out protocols consisting of anywhere from three to thirty contracts, then abstract a subset of functions from those contracts into a single frontend app with a defined sequence of transactions. However, these apps do not reveal all possible user paths that can be called with functions on the underlying contracts. There are many reasons for this - some functions are role-dependent, some are meant for power/technical users, and others are still being beta-tested. This is fine if a protocol is meant to exist independently, without composing with (calling) other contracts deployed by other teams. But that isn’t the case, as what we’ve built up in crypto today are many "hyperstructures” that are meant to be developed on top of. Hundreds of contracts work together to form this composable web of onchain actions. Let’s call a set of actions (transactions) across this web a “flow.”
What does an example flow look like? You could create a new token on a platform such as Virtuals, reward those who mint the bonding curve using Boost, make mint fee payable from any chain using Relay, launch it into a Uniswap pool, and reward LPs with a portion of your fee earnings using 0xSplits, save any passively earned ETH on Aave, and then let collectors run and vote in fun contests on Jokerace on how to spend the ETH.
Think of each protocol as a potential expansion of the interaction space of other protocols. Each protocol can be called in sequence with another protocol onchain given a set of directly compatible functions (or requiring one more contract in the middle). This means that there is the possibility for 2^n interactions for each new protocol deployed, where “n” represents a protocol. Newer protocols often allow developers to arbitrarily integrate other contracts through extensions. Seaport (opensea) developed zones for smarter returns on passive liquidity in pools, Safe developed account modules for more granular token and wallet management, and Uniswap developed hooks for enhanced liquidity and swap rules in transactions. All of these increase the number of potential flows between different protocols.
But if users can’t discover and use these flows, we will never unlock the long tail of possible interactions. Unfortunately, exploring this “flow” across contracts under the app layer is left to only devs and analysts to play with. The other 99% of the community gets simple apps that abstract only a subset of contract functions and are single-protocol focused. Composability exists onchain, but its value is locked away for the majority of users - leaving them to only create and swap tokens. At Herd, we believe this whole layer of “flows” can be unlocked if we make contracts accessible to everyone to search, understand, and act upon.
The Difficulty Behind Exploring Onchain
There are two core components to exploring onchain flows before you can act:
* Search: Find matching actions (functions on a contracts) given an user intent
* Understand: Understand what a set of functions do and their relationships
Block explorers today force you to navigate dozens of pages to find information and only really support searching for tokens. You can use platforms like Dune and DefiLlama to learn about the data behind various protocols, but they are still abstracted from the actual contracts themselves. In addition, most protocol contract documentation is outdated, lacking in depth, or missing altogether.
The difficulty that both beginners and experts currently face in searching for and understanding contracts is the first significant barrier we want to break down with Herd. Last October we released our first agent, Sage, in a Telegram bot that lets you search for contracts with natural language text. When you select the contract, our agents will bring other relevant contracts into the chat context. We’ve since expanded from just analyzing contracts to transaction and wallet analysis. We are incorporating feedback from our five hundred beta users into a more powerful chat-based coterminal, currently under development.
Join the Herd
The founders, Andrew Hong and Edward Bramanti, spent the last three years building at Dune on both the engineering and devrel teams. We’re deeply familiar with onchain data, smart contracts, and the crypto community. We’ve always felt that enormous value is hidden in the composability between protocols, and also seen firsthand how inaccessible that value is to almost everyone - even within top teams in crypto. And we’re ready to change that.
>> We're hiring engineers to join us (full-stack engineer and LLM founding engineer)
>> Sign up for the web app beta, releasing this spring.