Crypto trading involves substantial risk. POEOS does not guarantee profits. POEOS is research tooling: it reports on-chain data and never predicts prices or guarantees outcomes.

Grants & Research

Public record

Everything a reviewer needs to evaluate POEOS: what problem it solves, how it is built, what is genuinely working today and what is explicitly not.

Honest-by-default reporting

POEOS publishes real counters or nothing at all. There are no fabricated user numbers, no simulated volume presented as traction, and no profit claims anywhere in this product.
Registered accounts

—

Live database count

Token analyses saved

—

Read-only chain lookups

Simulations recorded

—

Paper trading only

Strategies configured

—

Paper mode

Loading platform counters…

Mission

Why POEOS exists

Retail and independent developers enter DeFi with almost no tooling for verification. Token risk is judged from social signals rather than on-chain facts, strategies are executed without ever being modelled, and the cost floor of a trade is discovered only after paying it. POEOS is infrastructure for the step *before* execution: read the chain, score what is observable, model the economics honestly, and make the limits of that analysis visible.

Problem

What is broken today

  • Token due diligence requires reading raw RPC output most users cannot interpret.
  • Arbitrage and flash-loan content promises profit while hiding gas, fees, price impact and failure cost.
  • Multi-chain work forces developers to stitch together incompatible EVM and SVM tooling.
  • Analytics products rarely disclose their data sources, staleness or blind spots.

Architecture

How the system is put together

Client

React + TanStack Router single app. No secrets in the bundle, no private keys, no seed phrase input anywhere in the codebase.

Server functions

All RPC, market-index and AI calls run server-side. Provider URLs and keys stay in server environment variables and are never returned to the browser.

Chain adapters

Separate SVM and EVM modules. Solana reads parsed mint accounts and largest holders; EVM reads ERC-20 calls and contract code. They are not pretended to be interchangeable.

Risk engine

Pure, dependency-free scoring module with published weights, per-factor observations and an explicit coverage percentage for unknown data.

Persistence

Postgres with row-level security on every user table. A user can only read and write rows matching their own auth id.

Simulation

Deterministic calculators for routes, fees, gas, price impact, slippage and flash-loan premiums. No execution path exists in the codebase.

Roadmap & measurable milestones

M1 — Read-only intelligence

In progress
  • Live SPL + ERC-20 token reader across Solana, Ethereum, Base and Arbitrum
  • Transparent, factor-by-factor risk signal with published weights
  • Real RPC/API health monitoring with latency and last-checked time

Success measure: A reviewer can paste any mainnet token address and reproduce every number from the displayed sources.

M2 — Simulation & persistence

Partly shipped
  • Strategy Lab and Flash Loan Lab as deterministic calculators with published formulas
  • Per-user saved analyses, paper runs and bot configurations behind row-level security
  • Shared Risk Controls panel with minimum-net-profit abort logic

Success measure: Every simulation is reproducible from its stored inputs; no run is presented as an executed trade.

M3 — Open developer tooling

Planned
  • Open-source ABI/JSON/RPC utilities and contract templates
  • Public SDK for the risk model so third parties can audit and reuse it
  • Published methodology document and test vectors

Success measure: Risk model released under an open licence with a reproducible test suite.

M4 — Reviewed execution boundary

Not started
  • Non-custodial execution behind explicit per-action user signing
  • Third-party security review before any mainnet execution path is enabled
  • Transaction simulation and revert-if-unprofitable enforced on-chain, not just in UI

Success measure: No execution feature ships without a published audit and an opt-in, per-transaction signature.

Public-good components

Intended for open release

  • Token risk model (weights, factors and test vectors)
  • EVM + SVM read-only token adapters
  • Flash-loan and arbitrage economics calculators
  • ABI/JSON/RPC developer utilities
  • Endpoint health probe library

Repository link is reserved here and will be published with the first tagged release.

Transparent limitations

  • POEOS does not execute trades. Nothing in the product places an order or moves funds.
  • The risk signal reads observable on-chain and indexed market facts only — not contract source, audits or team identity.
  • Public RPC endpoints are rate limited; dedicated providers are required for production reliability.
  • Liquidity and volume come from a third-party index and may be delayed or incomplete.
  • No POE token has been launched. Utility described in the product is a specification, not a live asset.
  • The platform has not undergone an external security audit.