How to Write Exit Rules With AI
This prompt turns a vague exit plan into a stop price, one or two take-profit prices, and a trailing stop, plus the dollar result if each rule fills at the price you wrote. No AI can predict markets. A written price is not a promise that the order fills there.
How to use this prompt
- Fill in the prompt Replace every [BRACKET] in the exit prompt with your own numbers. Leave a bracket blank only if you truly do not have that input.
- Run it in your AI Paste the plain prompt into any chat assistant. For the Moby version, use a supported host that is already connected with your Moby agent key.
- Check the answer against the checklist Each rule should name which part of the position it sells. The dollar column should match price times size, before fees. Two rules must not both try to sell the same tokens.
The prompt (any AI chat)
You paste the entry, the size and the plan in your own words. The assistant returns prices and sizes. It does not place anything.
Turn my exit plan into exact rules I can enter as orders.
Entry price: [PRICE] Position size: $[SIZE]
My plan in plain words: [e.g. "cut it if it drops a lot, take some profit if it runs, let the rest ride"]
Give me: a stop loss price, one or two take-profit prices with how much to sell at each, and an optional trailing stop (trail % and activation price).
For each rule show the dollar result if it fills exactly at that price, before fees.
Then list two ways these exits could fill worse than planned.
The same prompt through the Moby MCP
Agents place their own exits. Moby does not add one automatically. A trailing stop sells at market when the price falls by your trail percentage (0.5% to 50%) from its highest point, with an optional activation price. It is not available on SOL positions, and no fill price is promised. You can keep multiple resting orders. Exit orders on SOL positions are not supported for agents. No fill is promised.
Supported hosts: Codex, Claude Code (CLI), Gemini CLI (with a paid Gemini API key or enterprise access) and Grok CLI.
ChatGPT: works on Business, Enterprise and Edu workspaces that allow custom MCP servers (web).
I hold [TOKEN]. Using create_order, propose (do not place yet):
1) a stop loss at [X]% below my entry, 2) a take profit selling [Y]% of the position at +[Z]%, 3) a trailing stop with trail_pct [T] and activation_price [P] for the rest.
Check get_my_position first. Show me each order's details.
Before any trade, tell me the token, contract, amount and quoted price, and wait for my yes.
Worked example
Example, not a recommendation. Illustrative output, not real market data. A $50 position, three exit rules, three price paths. Entry price $0.0200, so 2,500 tokens before fees. Token counts ignore the entry fee for clarity. The $0.90 buy fee is shown separately.
A 0.75% swap fee (minimum $0.90 per swap), plus onchain fees. On a $50 market buy, 0.75% is $0.38, so the $0.90 minimum applies. Limit, take-profit, stop-loss and DCA fills are 0.75%. The stop and the take profit in this example are those fills, so their fee is 0.75% of the fill, not another $0.90. The table below is before fees, which is what the plain prompt asks for. A trailing stop is described by the sentence above. This page does not add a separate fee claim for it.
| Rule | Price | What it sells | Result before fees |
|---|---|---|---|
| Stop loss, down 20% | $0.0160 | All 2,500 tokens | Loss of $10.00 if filled at the trigger |
| Take profit, up 50% | $0.0300 | Half, 1,250 tokens | $37.50 back on $25.00 of cost, a gain of $12.50 |
| Trailing stop, 15%, arms at $0.0300 | 15% below the high after it arms | The other 1,250 tokens | Depends on the high, see the paths |
| Path | What happens | Result before fees |
|---|---|---|
| Falls to $0.0150 first | Stop triggers at $0.0160 | Loss of $10.00 if filled at the trigger, worse if it gaps |
| Runs to $0.0400, then falls | Take profit sells half at $0.0300. Trail sells the rest at $0.0340, which is 15% below the $0.0400 high | Gain of $12.50 on the first half and $17.50 on the second, $30.00 together |
| Sits at $0.0210 | Nothing triggers | Orders keep resting. Result is still open |
Path A: 2,500 tokens times the $0.004 drop is $10.00. A fill at $0.0150 is a $12.50 loss before fees. Path B: half at $0.0300 gains $12.50, and the trail at $0.0340 gains $17.50, together $30.00 before fees, only if the trail fills at that price. A trailing stop sells at market when the price falls by your trail percentage (0.5% to 50%) from its highest point, with an optional activation price. It is not available on SOL positions, and no fill price is promised. Path C sits at $0.0210 and hits nothing. You can keep multiple resting orders. If a take profit and a stop cover the same half, cancel the other when one fills, or ask your AI to.
More on the order types is on stop loss and take profit. A reward-to-risk check for the $10 loss against the $12.50 first target is on the risk reward calculator.
How to get a better answer
- Say which part of the position each rule covers, in tokens or in percent.
- Ask for the dollar outcome of each rule at the price you wrote, before fees.
- Ask the assistant to check that two rules do not sell the same tokens.
- Ask for the worst realistic fill, not only the trigger price.
What this prompt cannot do
- It cannot promise a fill at the trigger price.
- It cannot react while your AI client is off. Resting orders still work on Moby.
- It cannot set an exit on a SOL position through an agent.
FAQ
What are exit rules in trading?
They are the prices and sizes where you will sell, written before the trade so a mood does not pick the exit. A stop loss, a take profit and a trailing stop are three common ones. Each should say how much of the position it sells.
Should every trade have a stop loss?
A stop is how you name the loss you are willing to take. It is still your choice. A stop does not fill at the trigger if the price gaps, and it is not available on a SOL position through an agent.
How does a trailing stop work?
A trailing stop sells at market when the price falls by your trail percentage (0.5% to 50%) from its highest point, with an optional activation price. It is not available on SOL positions, and no fill price is promised.
Can my AI agent place a stop loss for me?
Yes, as an order it places after you confirm. Moby does not add exits automatically for agents. You ask for the stop, the agent proposes it, and you say yes before it is placed.
Why did my stop loss fill below my price?
The trigger is not a fill. If the next trade is below the stop, the sell can land there. The prompt asks for two ways the exit can be worse than the price you wrote. A gap is one of them.
Related prompts
- Prompt library
- Set trading limits
- Review a trade
- Explain a loss
- Stop loss and take profit
- Risk reward calculator
- Agentic trading
- Set up agentic trading
- AI trading bot
Try it with a Moby Agent
One agent per user, with its own wallet. Agents trade on Solana today. You set your agent's per-trade and daily spending limits and market-cap range; Moby blocks red-rated tokens. Limits are only as tight as you set them. Sells and exits skip the market-cap and safety checks, and tokens with no safety rating are not blocked. Turn on Ask before every trade and your AI asks you first. Moby does not hold or approve trades. AI Trading and copy trading currently need an access pass in the app. Download Moby.
In the Moby app, open the Agent tab and set up your Moby Agent. On the Connect your AI step, pick your platform and tap Create my key, then copy it. The key is shown once; if you lose it, rotate it in Agent settings. Say "Moby agent key" when you talk about it. Keep it out of the chat.
A 0.75% swap fee (minimum $0.90 per swap), plus onchain fees. Limit, take-profit, stop-loss and DCA fills are 0.75%.
The longer setup is how to set up agentic trading. How the loop works is on agentic trading. The map of bot types is the AI trading bot guide.