LarpBot

LarpBot is an open-source AI agent built by Tobiloba Adedeji (GitHub: tobySolutions) that targets one of the more persistent social problems in crypto communities: LARPing. In online crypto culture, "LARPing" (Live Action Role-Playing) refers to the habit of presenting fictional expertise, fake credentials, or unverifiable insider knowledge as though they are real — the anonymous "insider" who always knows what is about to pump, the pseudonymous "ex-institutional trader" whose calls conveniently never age well, the airdrop farmer claiming a team role. LarpBot is designed to detect these behavioral patterns in real time, focusing its analysis on X (Twitter) where the behavior is most prevalent.

What LarpBot Does

The bot functions as a conversational analytical tool rather than a blunt accusation engine. Its design philosophy centers on pattern recognition and neutral, objective analysis. When deployed and monitoring X, LarpBot identifies conversational signals associated with LARPing — inconsistencies in claimed expertise, unfalsifiable assertions, patterns of presenting speculation as knowledge — and responds through targeted questioning and constructive feedback rather than direct confrontation.

The goal is to surface and interrogate suspicious claims without triggering defensive reactions that typically shut down discourse. Users interacting with the bot receive analysis framed as information rather than accusation, which makes the tool more useful in community moderation contexts where preserving engagement matters.

Technical Architecture

LarpBot is built on the ElizaOS framework (formerly just "Eliza"), the TypeScript-based open-source AI agent operating system that became a foundational layer for the wave of autonomous agents deployed on Solana and across crypto in late 2024 and 2025. ElizaOS provides the scaffolding for persistent agent personalities, multi-platform connectors, and plugin extensibility that LarpBot relies on.

The agent uses Claude Vertex as its underlying language model for the actual analysis and response generation — giving it strong language reasoning capabilities for the nuanced task of evaluating conversational credibility. The project is Node.js-based, managed with pnpm and lerna for monorepo structure, and includes Docker containerization support for straightforward deployment.

For the X integration specifically, LarpBot uses ElizaOS's native Twitter/X connector, which handles real-time monitoring of posts, replies, and conversations. Installation follows the standard ElizaOS pattern: build with pnpm build, configure credentials and character files, and start the agent with pnpm start.

The developer also maintains a companion repository, larp-bot-plugin, which provides a plugin starter template for the ElizaOS ecosystem — evidence that the LarpBot work extends into contributing back to the broader agent infrastructure rather than remaining a standalone experiment.

Solana Connection

LarpBot's Solana integration is inherited through the ElizaOS framework rather than purpose-built. ElizaOS includes native Solana wallet support: private and public key management, Solana RPC endpoint configuration (defaulting to mainnet-beta), Helius API integration for enhanced on-chain queries, and SPL token trading capabilities via the Birdeye API. Any ElizaOS-based agent — including LarpBot — can be configured to interact with Solana programs, manage wallets, and execute transactions when those features are enabled in the character configuration.

Whether LarpBot's deployments actively use these Solana capabilities or rely on them passively as framework infrastructure is not specified in the public repository documentation. The project's primary stated focus is social media analysis rather than on-chain activity, but the wallet infrastructure is present.

LARP Token

There is an associated LARP token on Solana (contract address: EWQZj3fteHnkfad7oK46pES5QixJm7nbL8izM3xfzAHx), listed under the "Larpbot" name on Phantom. As of mid-2026, the token carries a market cap of approximately $2.8K, roughly $4.9K in liquidity, and around 231 holders. The token is flagged as unverified on Phantom, and the top ten holders control approximately 96.6% of the supply — indicating extreme concentration with minimal distribution.

The relationship between this token and the open-source GitHub project is not clearly established in any public documentation. The token appears to have been created as a speculative vehicle inspired by or loosely associated with the LarpBot concept, rather than representing any formal governance role, fee-sharing mechanism, or protocol utility. Investors should treat it as highly speculative given its market structure.

Developer Background

Tobiloba Adedeji is a Lagos-based senior software engineer with a substantive track record in both AI agent infrastructure and Solana development. He has worked with the Solana Foundation, Coinbase, Gaianet AI, and Livepeer, and is a core team member at Axis Protocol, a decentralized on-chain index fund (ETF) protocol on Solana. Other projects in his portfolio include Bourse (a Solana inference marketplace with USDC settlement), sol-fuzz (an AI-powered security scanner for Solana programs), and MCPGuard (a security layer for the Model Context Protocol). He has also contributed to open-source infrastructure projects including agentgateway and Million.js.

LarpBot sits in the experimental and developer-tooling segment of his work — an open-source contribution to the ElizaOS ecosystem that demonstrates practical application of the framework for social credibility analysis rather than a venture-scale commercial product.

Ecosystem Context

LarpBot entered the field during the ElizaOS agent proliferation of late 2024 and 2025, when Solana became the dominant platform for deploying autonomous AI agents. The wave produced hundreds of agents across trading, social monitoring, and community management use cases. LarpBot's niche — behavioral analysis of crypto social media for authenticity signals — addresses a real problem in the space, where anonymous account culture and hype cycles create fertile ground for the exact LARPing behaviors the bot targets.

The project is open source, publicly documented in both English and Spanish, and carries no known security audits given its nature as an agent framework deployment rather than an on-chain protocol managing user funds. Risks are primarily social and reputational: misidentification of legitimate commentary as LARPing, or use in coordinated harassment contexts, represent the meaningful edge cases for any deployed instance.

No venture backing, formal audit history, or team beyond the solo developer is documented.