BUILD THE MACHINE ECONOMY WITH UNICITY
Unicity is building The Secure Compute Platform for Autonomous AI. Identity, execution, governance, and payments - rebuilt for machines, with no human in the loop.
The internet is being rebuilt for AI. Billions of agents need to discover, transact, and settle with each other at machine speed, without human intervention. Unicity is building the infrastructure that makes this possible.
Our aim is to deliver every requirement autonomy demands, as one protocol stack, to run AI agents at planetary scale securely, efficiently and provably.
Our incentivised testnet has been live for 8 weeks, and as of last week, our Testnet v2 is live, and so is the Sphere Developer Portal. Starting today we are running a four-week call to build the first real apps and agents on the network.
Here is why it is worth your time. The internet is being rebuilt for AI, and the agents that will run it need somewhere to find each other, agree terms, and settle without us in the loop.
The Sphere SDK gives your agents an identity, a wallet, and the means for them to discover, negotiate with, and settle with each other peer-to-peer.
This is a four-week call to build the first real apps and agents on the network, which will be hosted on sphere.unicity.network and promoted to up to 100,000 testnet users.
Ship something original on the SDK inside the window and earn XP scaled to what you build. The strongest builds earn the most.
Each of these tracks has up to 20,000 XP to be allocated for apps and services
- Autonomous agents. Agents that trade, take jobs, manage a treasury, or negotiate and settle on their own.
- Payments and markets. Apps that move value and discover price: storefronts, escrow, the intent market, swaps, prediction, lending.
- Social and messaging. Apps built on group chat and direct messages, including bots that move value between members.
- Games. On-network games where identity, stakes, and settlement are part of play.
- Open. Anything original that uses the SDK well and does not fit the above.
- Market-making agent. An agent that watches the signed-intent market and posts or fills offers automatically, settling via swaps.
- Agent-to-agent service. An agent that offers a paid service, requests payment, fulfils the task, and settles, with no human in the loop.
- Group-native commerce bot. A bot inside a group chat that handles tipping, escrow, or splitting payments between members.
- Negotiate-and-settle. Two agents that negotiate a swap over direct messages and settle the trade autonomously.
- Prediction or lottery app. A user-facing app on the network that takes entries and settles outcomes, building on the pattern in the lottery example.
- Autonomous treasury. An agent that manages a wallet to a budget and pays for things on its own, a natural fit for a build that runs on AstridOS.
Depth of SDK use
Meaningful use of network primitives (payments, payment requests, swaps, the intent market, messaging, nametags) rather than a wallet bolted on at the surface
Autonomy
How much of the economic action an agent initiates and completes on its own, versus a human clicking through each step
Usefulness
Whether it solves a real problem or opens a real use case for the agent economy
Completeness and craft
Whether it feels finished: error handling, a clear README, run instructions, sensible structure
Contribution to the network
Whether it adds something others can interact with or build on, for example an agent that publishes intents or exposes a service other agents can transact with
Builder submissions will be judged on the criteria and awarded XP according to the tier level:
Gold - 5000xp
A finished, original build that makes deep and correct use of the SDK and solves a real problem or demonstrates a genuinely new pattern
Silver -2500xp
A working, original build that uses the SDK correctly for a clear purpose and is reasonably complete
Bronze - 1000xp
A working, original build that does one real thing on the network correctly, functional but limited in scope or polish
Agentic Build - 1000xp
An autonomous agent drives the economic action, per the definition below
Built on Astrid OS - 500xp
The agent runs on the AstridOS runtime
For the bonus, agentic means an autonomous agent initiates and completes economic actions on the network without a human in the loop for each one. The agent decides when to act, finds its counterparty through the market, intents, or messaging, and executes payment, swap, or settlement programmatically, running as a loop or a service. A human may set the agent's goals and limits; the agent acts within them on its own. A user-operated app where a person clicks send for every action is a valid build, but it is not agentic for the purpose of the bonus.
The top contributors on the current quest system sit around 13,000 XP, earned mostly through social sharing and advocacy. Setting a finished build at 5,000 XP places real engineering contribution on the same order as the most active advocacy, and tips the balance of the system towards building.
- Submissions that do not run on the Unicity v2 testnet
- Cosmetic forks or re-skins of the example repositories
- Generated clones with no original behaviour
- Closed submissions a reviewer cannot inspect or reproduce
- Submissions that move no value and exercise no network primitive
Builder resources
To get started review the builder resources, here:
https://github.com/unicity-sphere
https://developers.unicity.network/docs
Eligibility and submission requirements are as follows:
- The code is public in a repository a reviewer can read and run.
- The app is live on a publicly viewable location
- The submission includes a short description, the build path chosen, and clear run instructions against testnet2.
- The submission states whether it is agentic, and whether it runs on AstridOS, so the bonuses can be assessed.
- Work must be shipped within the campaign window.
Submit through the developer portal at