Say that custody is a journey, without arguing anyone into buying

[HodlDee]

What this changes, and why

Closes #65. New guide: Self-custody is a journey, not a test you pass (custody-journey, fundamentals, beginner).

The brief asked for the journey framing — day one on an ETF or an exchange is fine and should not be frowned upon, the goal might be multisig, and you do not drive on the motorway during your first lesson.

The thesis changed during review, and this description is the corrected one

The original draft argued that having exposure beats waiting. It called reading about custody for eight months and buying nothing "the genuinely common failure", ranked someone with coins on an exchange above someone with a good mental model and none, and pull-quoted "having exposure is not the mistake."

External review flagged that as persuasion about whether to own bitcoin — in the same batch where #116 was reframed precisely because the homepage says "Built to Inform, Not Persuade." Applying that standard to one guide and breaking it in another is worse than not applying it at all, so the thesis is gone.

What the guide argues now: whether to own bitcoin at all is not a question this site answers. Deciding it is not for you, or not yet, is a complete answer. What the page is about is custody — how to move between arrangements without hurting yourself, for people who have coins or have decided to get some.

Reducing shame about an existing exchange balance never required shaming someone who chose not to buy.

The first step needs no money, and now actually works that way

Replacing the pressure with an exercise that costs nothing was the right move, and my first attempt at it was wrong: it told the reader to "send a few thousand satoshis to yourself" while promising every mistake was free. An empty wallet cannot send anything, and a real transfer risks the amount and the fee. The reviewer caught it.

The unfunded exercise now ends where it can — set up a wallet, write the backup down properly, record the first receiving address, wipe, restore from the backup alone, compare the address. Every step is genuinely free because there is nothing in the wallet to lose, and the comparison is what makes it a test rather than a rehearsal of typing.

Sending is called out as the one thing it cannot cover, pointed at test-transaction, and marked explicitly optional and later.

It resolves a tension already in the library

choosing-your-first-setup says bitcoin left on an exchange "is exposed to the risk you were trying to avoid". True. The brief says day-one exposure is fine. Also true. The resolution is that the right response to that risk is a route out, not shame about being on it — and the two guides link to each other.

Notes for the reviewer

• The overlap was the risk. choosing-your-first-setup §5 already gives a three-stage arc. That guide answers which shape fits the amount you hold; this one answers how to move between shapes without hurting yourself. Cross-linked rather than repeated — worth checking you agree they are different questions.
• An ETF is price exposure rather than bitcoin — no withdrawal, because there is nothing to withdraw. Saying so is not shaming anyone for being in one.
• Readiness is presented as checkable rather than felt: restored once, sent and received deliberately, able to say where the keys are, and bored of the current setup. Boredom is the signal.
• It declines to assume everyone is heading for multisig. A hardware wallet with a tested backup is a complete answer for most people holding most amounts, not a waypoint. That is a position, and you may want it stated more or less strongly.

Tests

npm run build

Every guard passing. All glossary anchors resolve.

• I ran the tests relevant to this change
• npm run build completes with every guard passing
• If I added a test, I wired it into .github/workflows/build.yml

Generated output

• I changed build sources, and the regenerated docs/ is committed here

The offline artifact

• The digest is unchanged

Assets and third-party code

• This adds no images and no third-party code

Before requesting review

• This contains no real seed phrase, private key, passphrase or entropy
• This is not a security fix that should have gone through the private advisory form

🤖 Generated with Claude Code