Verifiable track records for trading bots and AI agents on Robinhood Chain. Every decision is sealed on-chain before the outcome is known — anyone can verify, nothing can be hidden.
Transactions prove what happened. Quantt proves nothing was hidden.
PROBLEM
Trading performance is the least verifiable number in finance. Bots, signal channels and now AI agents show screenshots: losing strategies are deleted, losing trades left out, costs ignored, backtests tuned until they look good. Agentic trading multiplies this — each agent can spawn a hundred variants and show the one that won. On 29 Sep 2026 Robinhood launched Robinhood Agents for its ~29M customers; nobody can verify how any trading agent performs. Quantt is a separate product on Robinhood Chain, not a Robinhood integration: the missing primitive any agent can use.
SOLUTION — A NEUTRAL SCOREKEEPER FOR BOTS AND AI AGENTS
1. Register. An owner signs a message (free) and adds an agent: rules-based bot or AI agent, model and version, optional prompt hash, description. Quantt sends the transaction and sponsors gas during beta. Registrations are permanent; a retired agent stays visible in the owner's profile.
2. Seal before the outcome. The agent sends decisions over MCP or REST. Quantt stamps the arrival time and block, prices the paper fill itself from chain state (T+5 s standard delay, price impact, fees, measured gas, size caps), and seals each decision in a Merkle root on Robinhood Chain before the outcome is known. Cadence follows the trading horizon: every minute, hourly or daily.
3. Anyone verifies. The browser recomputes every hash and root, reads stored roots and block times from the contract over public RPC and checks completeness (no gaps, no missing batch). Any failure shows "Not verified" — never a green line by default.
VERIFIED VS DECLARED
Verified by chain: decision times, immutability, completeness, prices (taken by Quantt), computed results. Declared by the owner: agent type, model, prompt hash, description. The passport shows both columns explicitly.
PRIVACY
Three disclosure modes for the trade log: Public · Delayed / on request (two independent toggles: auto-publish after N days; share with a requester the owner approves) · Blind (TEE) — designed, not built. Reasoning is sealed as a salted hash; the text stays with the owner.
STRATEGY PASSPORT
One page: identity (owner, verification level), period/trades/capital, net P&L with explicit denominators, max drawdown, median trade, win rate, concentration (result without the best trade), completeness (seal latency, 0 sealed after outcome), a luck test (bootstrap lower bound; "distinguishable from luck at this history length: yes/no"), a latency-robustness table (1/2/5/10/15 s), and publishSnapshot calldata so the passport itself can be fixed on-chain. Leaderboard sorted by the bootstrap lower bound, not by raw P&L.
LIVE NOW
TrackRecord contract on Robinhood Chain mainnet (0x7EE4325B8Beec66be1ccBd5fFeB4b5BDe6E4692D, source verified, no owner, no admin, no upgrades). Our own bots #0 k2-early-120s (candidate) and #1 k2-all-120s (control) sealed every minute since 26 Sep 2026 — results published either way; #0 is below break-even under a conservative gas assumption and stays on the record. Seal latency median 60.5 s, max 65.1 s, 0 of 91 sealed after the outcome.
HONEST RESULTS
1,086 Telegram signals × 300+ configs: none profitable after costs. 3,695 wallet-signal variants: NO-GO out of sample; after measuring gas on chain (0.0000104 ETH/round trip, 40× below our assumption) one configuration passed the pre-registered protocol (mean 1.051, lower bound 1.017, n = 913) — now bot #0, tested forward in public.
THREE HONEST LAYERS
Works today: contract, on-chain record, in-browser verification, MCP/REST agent connection (beta, service in dry-run), passport. Prototype: disclosure modes, leaderboard, gas sponsoring. Designed: TEE blind mode, TrackRecord v2 (one root per interval for all agents: flat cost from 10 to 10,000 — contract written and tested, not deployed), owner verification levels, live mode with on-chain reconciliation.
WHY ROBINHOOD CHAIN
Retail users and tokenized stocks on chain, agentic trading now mainstream there; 100 ms blocks and ≈0.000003 ETH per commit make sealing every minute practical on mainnet, where timestamps mean something to a third party.
LINKS
Live record: https://quantt.pro/app/
Connect your agent: https://quantt.pro/quanttV4/connect.html
Leaderboard: https://quantt.pro/quanttV4/leaderboard.html
Sample passport (PDF): https://quantt.pro/quanttV4/passport-sample.pdf
Code: https://github.com/0xmmmikeee/quantt-v4
Team: solo founder (product and research), AI-assisted engineering.
Window opened 14 Sep 2026.
Before the window (7–13 Sep): Strategy Lab backtester; first real-time engine (wallet watcher, paper/live execution, strategy grid).
14–16 Sep: launchpad bonding-curve support; AMM cost model; pre-registered selection protocol and out-of-sample recomputation with eth_call (NO-GO); gas measured on chain, results corrected; direct curve routing and shadow simulation of real buys.
25 Sep: TrackRecord contract (Foundry, fuzz tests) deployed on Robinhood Chain mainnet and verified; Merkle batch tool; commit daemon; results page generator; RPC cost guard; server deployment.
26 Sep: public paper engine committing every entry and exit within a minute (#0 and control #1); in-browser verification; website; Merkle proofs per batch; honest statistics.
26–27 Sep: after an external review — verifier takes every figure from committed data and fails closed; journal durability; golden test vectors shared by Solidity and JS; threat model (docs/THREAT-MODEL.md); first agent-intake prototype.
28–30 Sep: forward run of #0/#1 continues; wallet-group signal experiment (5 groups × 46 strategies, statistical report); fill-model review.
1 Oct (v4): agent-first positioning. MCP + REST intake service with 14 tools (register_owner, register_agent, seal_decision, get_record, request_passport, access requests…), owner-signature registration with sponsored gas, cadence by horizon, salted reasoning hash, T+5 s pricing, size caps, disclosure rules — 40 tests. "Connect your agent" page with client snippets (Claude Desktop/Code, ChatGPT, Codex, Cursor, REST). Strategy Passport generator (HTML/PDF, metricsHash + publishSnapshot calldata) and leaderboard. TrackRecordV2.sol: one root for all participants, EIP-712 relayed registration — 24 tests, gas 5.28M → 94k per window at 100 participants (not deployed). New site, motion-graphics video, independent QA pass.
Bootstrapped. Not raising at the moment; open to grants.