Verification on selfclaw is not avaialble due to reatrictions of self protocol verification in Nigeria. verification of project is verifiable on karma and 8004 with agent id 131
Osher Ai is an intent-based AI agent that enables users to transfer stablecoins from Celo to other blockchains fast, cheaply, and safely — without interacting directly with a bridge. The agent understands natural language commands, automatically detects destination chains, compares cross-chain routes in real time, selects the optimal bridge, and executes the transfer on behalf of the user. It also integrates price alerts and automated trading logic, ensuring users get the best possible execution before bridging. This project makes cross-chain transfers as simple as sending a message.
Moving stablecoins from Celo to networks like Base, Solana, Ethereum, Arbitrum, or Polygon today requires users to: Choose a bridge manually Compare fees across platforms Check liquidity Connect multiple wallets Approve complex transactions Risk sending assets to unsupported addresses This process is error-prone, expensive, and inaccessible for most users.
Osher Ai replaces manual bridging with an AI-powered autonomous agent that: Understands user intent in plain language Identifies destination wallet chains automatically Finds the cheapest + fastest + safest cross-chain route Validates addresses and asset compatibility before executing Provides clear feedback when a transfer cannot proceed Executes swaps when necessary to minimize costs Monitors prices and fees continuously Supports conditional, automated transfers Users simply say what they want — the agent does the rest.
- Intent-Based Transfers Users can say things like: “Send 100 USDT from Celo to this Solana wallet: 7xB2…” “Move 250 USDC from Celo to Base in the cheapest way.” The agent: Parses intent Detects the destination chain Selects the optimal bridge Executes the transfer
- Smart Cross-Chain Routing The agent evaluates multiple bridging providers, including: Wormhole LayerZero Axelar Hyperlane Portal Bridge Celer Across It optimizes for: Lowest fees Fastest delivery Highest success probability
- Built-In Safety Guardrails Before executing any transaction, the agent checks: Is the destination address valid? Does the chain support this token? Is liquidity sufficient? Are fees reasonable? If something is wrong, the agent explains clearly what needs to change (token, address, or route).
- Price Alerts & Auto-Trading Users can set conditions such as: “Alert me if bridging fees to Base drop below $1.” “Move my USDC to Solana only when gas is cheap.” “Swap my USDm to USDC before bridging if it saves money.” The agent can: Monitor prices across chains Compare cross-chain spreads Auto-swap assets before bridging Execute transfers only when conditions are met Example automated flow: Swap USDm → USDC on Celo Bridge USDC to Solana Convert USDC → USDT on Solana Deliver funds to user All autonomously.
Agent Layer Framework: JS + Claude compatible intent agent framework Capabilities: Natural language understanding Wallet detection Bridge selection Transaction orchestration Bridging Layer Connectors to multiple cross-chain protocols Real-time fee + liquidity comparison DEX Layer Uniswap Mento Curve Used for: Pre-bridge swaps Post-bridge conversions Data Layer Price oracles Gas monitors Liquidity trackers Wallet Layer WalletConnect / MiniPay Multi-chain support
Celo: USDm, USDC, USDT Destination chains: Solana Base Ethereum Arbitrum Polygon 🧠 Example Use Cases Use Case 1 — Simple Transfer User: “Send 200 USDT to this Base wallet: 0xA12…” Agent: Detects Base Finds best bridge Executes transfer Returns transaction link Use Case 2 — Conditional Transfer User: “Move my stablecoins to Solana only if fees drop below $0.80.” Agent: Monitors fees Executes automatically when condition is met
Phase 1 Basic intent parsing Wallet detection Single-bridge integration Phase 2 Multi-bridge routing Safety validation layer Price monitoring Phase 3 Auto-trading before bridging Conditional execution Dashboard UI Phase 4 Support for more chains Reputation tracking Performance analytics
Contributions are welcome! Please open issues or submit PRs with clear descriptions.
MIT