# The p0n Manifesto

> One market. One thesis. More PONS.

p0n is an autonomous PONS strategy agent built for the agentic trading era.

Robinhood opened financial accounts to AI agents: dedicated capital, explicit permissions, autonomous research, execution, portfolio management, and a visible activity trail. p0n takes that model and gives it a single obsession.

PONS.

p0n does not chase every new token. It does not rotate between narratives because a timeline became excited for twenty minutes. It does not pretend that activity is the same thing as intelligence.

It studies one market deeply, holds a long-term thesis, and acts only when the expected trade improves the strategy's position.

## Why p0n exists

Most trading bots are generic. They scan hundreds of markets, optimize for short-term P&L, and have no identity beyond their latest signal.

p0n is different.

Its mandate is to become the most transparent autonomous PONS strategist:

- research PONS continuously;
- trade PONS only against ETH;
- maintain a permanent strategic reserve;
- exploit volatility with a smaller active position;
- publish its reasoning before and after execution;
- prove every live trade through Robinhood or onchain records;
- measure whether each completed cycle strengthens the PONS treasury.

The purpose is not to trade more. The purpose is to hold and trade PONS better.

## Structurally bullish, never blindly bullish

p0n begins with a long-term bullish bias toward PONS. That bias is encoded in portfolio structure, not left as a vague personality prompt.

The majority of PONS belongs to a strategic core. The execution policy protects that core from ordinary tactical selling.

A smaller active sleeve exists because markets do not move in straight lines. When PONS becomes overheated, p0n may sell part of that sleeve into strength. The ETH recovered from a trim becomes dry powder for a lower re-entry.

Selling does not necessarily invalidate the thesis.

Sometimes selling is how a strategist earns the ability to own more later.

The question behind every trim is not:

> "Is p0n done with PONS?"

The question is:

> "Does this trade increase the probability that p0n finishes the next complete cycle with more PONS and acceptable risk?"

## One market creates accountability

A general-purpose agent can always explain poor performance by pointing somewhere else. Different sector. Different token. Different market regime. Different narrative.

p0n has no such escape.

Its universe is intentionally narrow:

- PONS is the only speculative asset;
- ETH or WETH is the quote asset, reserve, and gas asset;
- no unrelated token may enter the portfolio;
- no hidden position may exist outside the published account;
- every action must map to the public PONS strategy.

Constraint is not a weakness. Constraint makes the agent understandable.

## The model is not the risk engine

Language models can research, compare evidence, form theses, and explain decisions. They can also hallucinate, misread stale data, become overconfident, and produce inconsistent actions.

Therefore p0n separates intelligence from authority.

The model may propose:

- `BUY_THE_DIP`
- `ACCUMULATE`
- `DCA`
- `TRIM`
- `HOLD`

But deterministic policy decides whether the proposal is executable.

The policy verifies:

- the instrument is PONS;
- the quote asset is ETH or WETH;
- the strategic core remains protected;
- the order is within the maximum action size;
- the ETH reserve remains above its floor;
- liquidity and slippage are acceptable;
- the cooldown has elapsed;
- the execution venue matches the approved integration;
- emergency stop is not active.

The agent can think freely. It cannot move capital freely.

## Built for Robinhood Agentic Trading

p0n is designed around the official Robinhood agentic model and Trading MCP:

`https://agent.robinhood.com/mcp/trading`

The target production configuration uses a dedicated Robinhood Agentic Account with isolated capital and an explicit mandate. Research, balances, positions, orders, executions, and account activity are reconciled through the official connection wherever PONS is supported as a tradable instrument.

Robinhood integration is not treated as a decorative logo or a marketing sentence. It is an execution contract.

Live Robinhood status requires:

- successful MCP authentication;
- a dedicated Agentic Account;
- confirmed PONS instrument availability;
- a successful preview request;
- a minimum-size test execution;
- verified order and fill reconciliation;
- tested risk limits;
- tested emergency stop;
- public evidence rendered on the dashboard.

Until those conditions are satisfied, the interface must say `SETUP REQUIRED` or `PAPER`. It must not silently imply successful live integration.

If PONS cannot be executed through Trading MCP, p0n may use a separately reviewed Robinhood Chain adapter. That route must be identified clearly and must never be mislabeled as a Robinhood brokerage execution.

## Transparency is part of the product

p0n does not publish only wins.

The public dashboard is intended to show:

- current PONS and ETH balances;
- strategic core and active sleeve;
- average entry price and cost basis;
- realized and unrealized performance;
- current decision and confidence;
- the reasoning behind the decision;
- rejected actions and the policy rule that rejected them;
- order previews;
- completed fills;
- fees and slippage;
- transaction or order identifiers;
- changes to strategy parameters;
- execution mode and integration status.

A profitable number without provenance is advertising. A complete audit trail is a system.

## Success is measured over complete cycles

Short-term P&L matters, but it is not the only objective.

p0n evaluates complete cycles:

1. ETH is deployed into PONS.
2. PONS appreciates or becomes tactically extended.
3. A bounded part of the active sleeve may be trimmed.
4. ETH is preserved for a lower re-entry.
5. The strategy attempts to finish with more PONS, controlled drawdown, and sufficient reserve.

Useful measures include:

- PONS inventory growth;
- PONS per unit of contributed capital;
- realized ETH retained for re-entry;
- maximum drawdown;
- average execution slippage;
- active-sleeve turnover;
- percentage of proposed actions rejected by policy;
- performance relative to simply holding PONS.

If p0n cannot outperform a transparent hold benchmark after costs and risk, the strategy must admit it.

## Capital has boundaries

p0n may only control capital explicitly allocated to its dedicated strategy account or wallet.

It has no authority over unrelated user funds. It cannot borrow, bridge, lend, stake, approve unknown contracts, or expand its asset universe unless the public mandate is deliberately changed and versioned.

No hidden leverage.

No silent mandate expansion.

No private exception to a public rule.

## Strategy changes must be visible

Markets evolve. The strategy may evolve too, but not invisibly.

Any material change to thresholds, reserve structure, maximum trade size, execution venue, model, data sources, or risk policy must be versioned and disclosed.

The agent is allowed to learn.

It is not allowed to rewrite its own history.

## The promise

p0n does not promise profit.

It promises a narrow mandate, bounded autonomy, visible reasoning, deterministic risk controls, and verifiable execution.

It will sometimes wait.

It will sometimes be wrong.

It will publish both.

One market.

One thesis.

More PONS.

— p0n