Step 1: a wallet that exists only for this
Create a new Solana wallet used for nothing else. Not the one holding your other positions, not the one connected to anything you care about. This takes two minutes and is the single highest-value thing on the list.
The reason is not distrust of any particular tool. A trading wallet interacts with brand new token contracts that by definition nobody has vetted, and it accumulates relationships with them over time. Keeping that exposure in its own container is simply how you stop one bad contract from reaching everything you own.
Step 2: decide the number you can lose
Pick the amount before you pick the tool, and pick it as the number you would be annoyed but not harmed by losing entirely. Then fund only that, and keep it separate from a fee buffer.
| Component | What it is for | Why separate |
|---|---|---|
| Position capital | The trades themselves | This is the part at risk |
| Fee buffer | Priority fees, tips, failed attempts | Spending it makes the next entry fail |
| Rent reserve | Token accounts being created | Small, but a hard requirement |
| Everything else | Not in this wallet | The whole point of step one |
A wallet funded to exactly the size of the intended position does not have enough, because fees are charged on top and new token accounts have to be funded. This catches almost everyone once.
Step 3: learn the venue before the tool
Sniping a bonding curve and sniping an AMM pool are different trades with different risks, and launchpads differ from each other more than most comparisons suggest. Graduation thresholds, fee structures and what happens at handover vary by venue.
Spend an hour reading how the venue you intend to trade actually works before configuring anything. It changes sizing, it changes exits, and it is the difference between a setting you chose and a setting you copied.
Venue by venue
- Pump.fun sniper botthe original curve and what changes at graduation
- LetsBonk sniper botLaunchLab rails, graduating near 85 SOL
- Boop.fun sniper bota 400 SOL curve and a longer screening window
- Believe sniper botthe venue that charges you for arriving first
Step 4: choose the tool
Only now, and on four questions rather than on marketing.
- 1Who holds the keyCan you export it. If not, the service holds your funds regardless of how the interface describes it.
- 2How does it learn about launchesDecoding program logs in-slot, or polling an aggregated feed. The second cannot be fixed by anything downstream.
- 3Which venues does it actually decodeA logo grid is not coverage. Ask which programs are decoded, and what happens when a new launchpad appears.
- 4Where are exits enforcedIn the execution layer, or in the page you have open. Only the first survives closing the tab.
Step 5: write the screening profile
Decide which risks you will not take, in writing, before anything is armed. The default answer for a first week should be strict, because the cost of missing a winner is smaller than the cost of learning what a honeypot feels like.
- Mint authority revoked, so supply cannot be inflated after you buy.
- Freeze authority revoked, so your ability to sell cannot be switched off.
- Liquidity burned or locked, so the pool cannot simply be withdrawn.
- A simulated sell that succeeds, which is the cheapest honeypot test available.
- Limits on developer holdings and on how concentrated the early distribution is.
Every one of these removes candidates, including some that would have worked. That is the trade you are consciously making, and it is the right trade while you are still learning what your own fill rate looks like.
Step 6: write the exits before the first entry
Take profit levels, a stop, trailing behaviour and a maximum hold. Attached automatically when the position opens, enforced by the execution layer, not by you watching.
This step is the one most people skip, and skipping it is what turns a bad entry into a bad month. An entry decides what you paid; an exit decides what you keep.
| Control | Sensible starting posture for week one |
|---|---|
| Take profit | Laddered, so no single guess has to be right |
| Stop loss | Present, and wide enough to survive ordinary noise |
| Trailing | On, tight enough to protect a gain that happened |
| Maximum hold | Short, because early positions are for learning |
Step 7: spend a week measuring
The first week is not for making money. It is for producing four numbers you do not currently have, and which no amount of reading can give you because they depend on your settings, your venues and your sizing.
- 1Fill rateAttempts divided into fills. This is the denominator in every cost calculation you will make afterwards.
- 2Average gap between quoted and actualYour real slippage, as opposed to your slippage setting.
- 3Cost per filled positionAll fees including failed attempts, divided by positions actually opened.
- 4Expectancy per attemptTotal result divided by attempts. If this is not positive, nothing else in the setup matters yet.
Questions people ask
These come up in the ticket queue often enough to answer once, properly. If yours is not here, send it over and it probably ends up on this page.
Create a dedicated wallet, decide a funding amount you can lose entirely, learn how your target venue works, then choose a tool on custody and detection rather than marketing. Write a screening profile and exit rules before the first entry, and spend the first week measuring four numbers rather than trying to profit.
Key takeaways
- Choosing a tool is step four, not step one.
- A dedicated wallet takes two minutes and caps the worst case at what you put in it.
- Fund position capital and a fee buffer separately; a wallet funded to exactly the position size will fail.
- Learn the venue before configuring, because thresholds and handovers differ by launchpad.
- Write screening and exits before the first entry, and have exits enforced by the execution layer.
- Week one is for producing fill rate, real slippage, cost per position and expectancy per attempt.