buff.trade

buff.trade was a Solana-based platform developed by the team behind Homo Memetus that allowed anyone to tokenize an AI-powered trading strategy and launch it as a revenue-backed token. The platform core premise was that trading performance, not speculation, should determine token value. The project was founded in December 2024 and reached a series of early milestones before its GitHub organization was archived as "no longer maintained" in June 2026, indicating the project is no longer active.

What buff.trade Was

buff.trade positioned itself as infrastructure for an "Internet Capital Market" centered on verifiable trading performance. Where most crypto tokens derive value from narrative or community sentiment, buff.trade tokens were designed to represent a fractional ownership stake in a live AI trading agent and a direct claim on the revenue that agent generated from managing capital.

The platform tagline captured the intent concisely: "Tokenize trading strategies powered by trading agents." Users could submit a trading concept, which the platform would package into an AI agent that would then execute that strategy autonomously. The resulting token gave holders exposure to actual returns of the underlying strategy rather than a speculative bet on future adoption. According to the project own documentation, the goal was to create "a fair, efficient, and accountable Internet Capital Market centered on verifiable trading performance."

How the Platform Worked

buff.trade operated in three sequential stages: strategy submission, performance validation, and on-chain fund management.

Strategy submission and tokenization. A user submitted their trading idea, which the platform converted into an AI trading agent and wrapped in a tradeable Solana token. The token represented ownership in that specific agent and its associated strategy. Tokens were not immediately managing real capital — they first entered a simulated competition environment to prove merit.

Virtual competition and performance validation. AI agents competed in a transparent simulation environment called the Survival Dashboard. This served as a merit filter: agents had to demonstrate real returns before gaining access to investor capital. The platform documentation described this as requiring tokens to "demonstrate real returns in virtual trading competitions before managing actual capital." The Survival Dashboard launched in April 2025. By the time of the project September 2025 accelerator cohort acceptance, the team reported that more than ten agents had achieved over 100% profit-and-loss returns within two months of the dashboard going live.

Token graduation and fund management. Tokens that reached a threshold of 85 SOL in bonding curve progress graduated from the simulation phase. Upon graduation, the token migrated to a Meteora liquidity pool, and the underlying strategy became eligible for the on-chain fund management phase — where the AI agent would begin trading with actual capital and distributing revenue to token holders. During the prototype period, fund management operated in a simulated capacity, with live capital management planned for a subsequent public launch.

Revenue sharing was a core design element: token holders received distributions from profits generated by successful strategy execution, creating direct alignment between token value and fund performance rather than speculative market dynamics.

Ecosystem and Token

The broader ecosystem was anchored by the HOMO token (contract address: 7Uuzh9JwqF8z3u6MWpQuQJbpD1u46xPDY6PGjwfwTh4o on Solana), the native token of the Homo Memetus project. The buff.trade platform was built as the primary product expression of Homo Memetus, with HOMO serving governance and utility functions within the ecosystem.

The team described the project broader ambition as building the world first agent capital market — where AI trading agents compete, evolve, and launch as tokenized, revenue-sharing funds based on real performance, not hype. This framing distinguished buff.trade from contemporaneous AI agent token launches that primarily generated speculative trading activity without a concrete link to verified strategy execution or returns.

Technical Foundation

buff.trade was built on Solana and used smart contract automation throughout its pipeline — from token creation and simulation through profit distribution. The team released a series of Solana-focused developer tools under the memetus GitHub organization, including a Pump.fun token monitoring SDK written in TypeScript, a rug analyzer, an agent training module in Python, an agent tokenizer written in Rust, and both server and client implementations using the Model Context Protocol (MCP). The platform documentation also referenced open-source contributions and ecosystem integrations as part of its technical architecture.

Recognition and Partnerships

The buff.trade and Homo Memetus team won the OKX Solana Accelerate Hackathon, a developer competition targeting DeFi and AI applications using the OKX DEX API on Solana. The project was subsequently selected for the OnePiece Labs x Solana Accelerator cohort that launched in September 2025. Reported partnerships at that time included integrations with elfa.ai, a social intelligence platform, and TGmetrics, a Telegram analytics tool. The team was also in the process of raising a $500,000 SAFE plus token warrant round as of September 2025.

Current Status

The memetus GitHub organization — which housed all buff.trade and Homo Memetus open-source code — was archived on June 7, 2026, with an explicit note that it is "no longer maintained." The buff.trade repository within that organization was last updated on October 30, 2025. The buff.trade website is no longer accessible. The @homo_memetus Twitter account has shown no recent activity. The HOMO token continues to trade on decentralized exchanges including Raydium and Jupiter at a substantially reduced market capitalization compared to earlier periods.

buff.trade represented a substantive attempt to solve a persistent problem in the AI agent token space — the absence of a verifiable link between token value and actual trading performance — and achieved a series of early technical and competitive milestones before being discontinued.