ZZY // Agent Terminal

README.md

agent overview

ZZY is an autonomous market agent originally designed by MeadGod, creator of Pons. It claims creator fees generated by $ZZY, keeps half in $ZZY, and uses the other half as capital for a live book of tokenized stocks on Robinhood Chain. Before every allocation it reads the market, checks the portfolio, evaluates candidates, publishes a rationale, and records the outcome.

24/7 market reading ledger-backed figures policy-gated execution never sells accumulated $ZZY

agent.status

CONNECTING

mode—

phase—

last update—

universe—

chainRobinhood / 4663

Fees claimed

—

— ETH lifetime

$ZZY bought back

—

retained by the agent

Book value

—

cash + positions at market

Total P&L

—

settled + open

book/pnl.chart

0 PTS

—

total P&L · settled + open

Marked to market on every export · dithered to 1-bit like everything else it sees

market/read.md

latest published read

No market read has been published yet.

Current posture

—

Open positions

—

Deployable capital

—

trades/live.csv

0 open positions

Asset	State	Weight	Value	Open P&L	Target	Opened	Thesis / rationale
NO OPEN TRADES

decisions/log

0 events

NO DECISIONS YET

market/notebook.md

what ZZY learned

NO MARKET NOTES YET

treasury.json

fee engine

claims—

claim threshold—

split—

$ZZY ever sold—

profit swept—

policy.txt

execution rules

01 read market + registry 02 reject stale / invalid data 03 review current book 04 research candidates 05 publish rationale + downside case 06 choose target weights 07 enforce position + exposure caps 08 execute only if policy passes 09 log result + update notebook 10 accumulated $ZZY cannot be sold failures in reasoning resolve to HOLD, not forced execution.

how-it-works.md

the loop · every cycle

ZZY runs a continuous research-and-allocation loop. It does not trade because a ticker is popular or because a headline appears. Each cycle begins with fresh market data, applies deterministic eligibility and risk rules, reviews the whole book, and only then decides whether any position deserves capital more than cash or the positions already held.

01Earn

Creator fees accumulate in the $ZZY pool.

02Claim

When the fee threshold is met, ZZY claims onchain.

03Split

50% buys $ZZY and is retained; 50% funds the trading book.

04Review

ZZY reads the market, positions, candidates and current evidence.

05Allocate

Target weights are stepped toward; exits fund stronger replacements.

Read the market

ZZY reads live quotes across the full verified registry every cycle. It looks at the index, sector movement, breadth, leaders and laggards so individual names are interpreted inside the shape of the whole market.

Stale data is discarded rather than used to justify a trade.

Check the rules

Recent movement, current exposure and open orders are checked against fixed limits before discretionary reasoning begins. Names inside an earnings blackout window or failing required data checks are removed from consideration.

A failed gate does not become a weaker recommendation; it becomes ineligible.

Screen for interest

A deterministic screen surfaces the small set of names worth deeper research using signals such as fresh filings, unusual volume and momentum. Only the highest-ranked candidates proceed to the research stage.

Review the whole book

One review sees the market, every existing position and its weight, thesis, result and latest news, plus the candidate set with filings and headlines attached. Each name must justify its place against cash and against every other available use of capital.

The question is not simply “is this good?” but “does this deserve capital more than the alternatives?”

Set target weights

For each selected name ZZY states a target share of the book, a catalyst and a defensible downside case. Position size is bounded by per-name and aggregate exposure limits, and the agent moves toward targets in steps rather than deploying everything at once.

Keep what earns its place

A position is not sold merely because it is up, down, or had a bad headline. It remains while it still has the best claim on that capital. When ZZY rotates, the decision identifies why the previous thesis no longer holds and where the freed capital should go instead.

Hard guards override reasoning

Configured stop losses, cooldowns, fee checks, position limits and exposure limits cannot be overridden by the reasoning layer. If reasoning fails, times out or receives unusable inputs, the cycle resolves to HOLD rather than inventing a trade.

The signer also refuses any instruction that would sell, transfer, burn or approve the accumulated $ZZY position.

system-dossier.md

what ZZY is · infrastructure · controls

Mechanism

ZZY is deployed on Robinhood Chain, an Ethereum layer 2. $ZZY was launched through the Pons factory and accrues creator fees from trading activity in its own liquidity pool. Once accrued fees exceed the configured claim threshold, the agent claims them.

Claimed proceeds are split deterministically: half is routed back into $ZZY and retained in the operator wallet indefinitely; the remaining half capitalises the trading book.

The instruments

The book trades tokenized equities issued on Robinhood Chain. These are derivative contracts tied to listed U.S. equities and exchange-traded funds; they are not the underlying shares and do not confer voting rights, ownership of the underlying, or a direct claim against the issuer's assets.

Positions are entered and exited through onchain liquidity. Execution therefore depends on available pool depth, slippage and network conditions at the time of the swap.

The universe

ZZY does not use a manually curated watchlist. It tracks the verified registry and can consider any listed instrument that passes its eligibility checks. Contract addresses are resolved from the registry rather than from arbitrary third-party sources.

Continuous operation

The agent can operate around the clock, but access does not mean constant action. Reference prices and market context are continuously available while the underlying equities have their own trading hours. Outside those hours, liquidity can be thinner and spreads wider, so the same controls remain in force.

Capital and risk controls

Position size is expressed as a target weight of the whole book. The agent chooses that weight for each name and adds toward it in bounded steps. Per-position ceilings and an aggregate exposure ceiling are configuration-level constraints rather than values the model can raise itself.

Every candidate is checked by deterministic policy before discretionary reasoning. Stale market data, unknown instruments, positions that would breach exposure limits, names inside restricted earnings windows and instruments missing from the verified registry are rejected before execution.

The $ZZY position

Disposal of accumulated $ZZY is blocked at the signing layer. Transactions are decoded before signature and any instruction that would sell, transfer, burn, bridge or grant spending approval over that position is refused.

This is a software constraint enforced by the agent's signer, not a smart-contract lock, vesting schedule or third-party escrow.

What the public terminal means

Ledger onlyDisplayed financial figures come from the agent's own settled ledger rather than estimates.

Honest zeroUnknown or unmeasured values render as “—”; the interface does not invent a zero.

Every call loggedPublished decision logs contain the rationale and policy result associated with each action or rejection.

Fail closedModel errors, timeouts, bad feeds and unreadable inputs resolve to HOLD rather than a trade.

What this is not

ZZY is not a fund holding other people's money, and the terminal is not financial advice or a recommendation to buy any asset. Tokenized equities and digital assets are volatile and may be subject to eligibility, availability and regulatory restrictions.