What the Migration Desk does
Migration Desk is a small editorial desk writing about one event: the moment a launchpad token stops trading on a bonding curve and starts trading in a pool. This page states what it covers, where claims come from, what it refuses to publish, and who it is not affiliated with.
Independence
This is an independent publication. It is not affiliated with, endorsed by, operated by or connected to Pump.fun, Raydium, PumpSwap or any other launchpad, exchange or protocol described on this site. Those names appear because they are the subjects of the reporting, in the same way any publication names the things it writes about.
No third-party logos, brand marks or product imagery appear anywhere on this site, and none of the design is intended to suggest an association that does not exist. If any page here reads as though it speaks for a protocol, that is a defect and the desk would like to hear about it.
What this desk covers
One subject, examined from three positions. The state a token is in while it trades on a bonding curve. The handover event itself, meaning the instruction that freezes the curve, creates a pool and moves two balances into it. And the state a token is in afterwards, where price is a ratio of reserves and every surface has to rediscover the venue.
What is out of scope is anything requiring a prediction. There are no price targets here, no assessments of whether a token is worth buying, no survival rates and no claims about what a chart does after graduation. The desk writes about a mechanism, which is a much smaller subject and one that can actually be described accurately.
Who writes it
Everything is published under one byline, The Migration Desk. There is no invented author, no fabricated biography, no stock photograph and no list of credentials the desk does not hold. A desk byline is an honest description of how the work is produced, and it avoids the alternative that is common in this market, which is a person who does not exist.
Why no parameter is printed as a number
Launchpad thresholds, migration fees, retained shares and timing behaviour are parameters set by whoever operates the launchpad. They live in program configuration or in a deployed program version. They have been changed before, and changes are not accompanied by a notification to every site that quoted the old value.
What follows is predictable. A number is published, other sites copy it, it becomes the number everybody repeats, and later people are budgeting against a figure that quietly stopped being true. The failure is silent, which is what makes it worth avoiding. Naming a parameter and explaining what it governs stays correct indefinitely; printing its value stays correct until somebody edits a config.
The same applies to graduation destinations. Where a launchpad routes graduating tokens is a live setting rather than a permanent fact, and it has changed across this ecosystem more than once. Every page here tells you to verify it against the protocol at the time of reading, and means it.
Documented, observed, inferred
Every claim about program behaviour belongs in one of three categories, and the category is stated where the claim appears rather than buried in a disclaimer at the bottom of a page.
- Documented. The project published it. Quotable, checkable, and attributable to its source rather than to a summary of it.
- Observed. Anyone can open a transaction or an account and see it. Reliable as description, weaker as explanation.
- Inferred. The desk worked it out and could be wrong. Always labelled, never allowed to harden into a stated fact.
The reason for the discipline is that unlabelled inference propagates. One guess about a threshold becomes a figure in a guide, then a requirement in somebody checklist, then a line in a budget, and by then nobody remembers that no protocol ever confirmed it.
Numbers
Numbers appear here in three forms only. Protocol facts, stated as facts. Arithmetic explicitly labelled illustrative, which uses round invented figures to demonstrate a mechanism and describes no real token. And figures attributed to a named public source a reader can open for themselves.
There are no statistics on this site about how launches perform, no graduation rates, no survival curves and no audience estimates, because the desk holds no dataset that would support any of them. Where a number would be useful and cannot be sourced, the page says so instead of supplying one.
What this desk will not publish
No fabricated statistics. No review counts, star ratings or aggregate scores. No testimonials, invented case studies or reconstructed profit and loss. No backdated publication dates. No screenshots presented as evidence of something they do not show.
Nothing that works by causing a reader to believe something false, regardless of how ordinary the practice has become. That includes implied endorsements from parties who expressed none, manufactured urgency, and any suggestion that this site speaks for a protocol it merely writes about.
And no promise of outcomes. Nothing here claims that following a procedure produces a result in the market, because a graduation is a change of venue and a change of venue entitles nobody to any claim about what happens next.
Commercial links
Some pages link to a commercial console for producing market activity on Solana pairs. Those links carry anchor text describing what they are, they open in a new tab, and they appear only inside passages that are genuinely about the mechanism they relate to. The desk does not place a link into a paragraph that was not already about the thing it points at.
The desk's position on that category is stated the same way everywhere on this site: producing activity changes what an activity-ordered screen displays, and it does not change whether anybody who sees the row decides the token is worth their attention. Both halves of that sentence always appear together, because dropping either one would misdescribe the product.
Corrections
If something here is wrong, the desk would rather know. Corrections are made to the page itself, and where a change alters the meaning of a claim rather than fixing a typo, the correction is described on the page rather than applied silently.
The most likely errors are in the inferred category, which is exactly why inference is labelled. A reader who can see which parts are worked out rather than confirmed knows which parts to argue with, and that is a more useful document than one presenting everything with identical confidence.
Contact
Questions, corrections and disagreements are welcome through the contact page. The desk does not accept payment for coverage, does not publish submitted articles, and does not write about a token because somebody asked it to.