docs(drift): draft from internal source

[pedromcunha]

Drafted from an internal source repo, requested by @arcticfloyd1984.

This PR was auto-generated by the reusable-docs-draft-pr workflow from a Claude-drafted update in the source repo. It is opened as a draft — please review, edit as needed, and mark ready for review before merging.

Review checklist

• Content matches the behavior shipped in the source PR
• No internal-only details (service names, file paths, unreleased names, security internals) leaked into the draft
• Conforms to CLAUDE.md style rules
• Cross-references and internal links resolve
• Any Needs human input items from the source comment are addressed

---

Generated by reusable-docs-draft-pr.yml in relayprotocol/agents-config.

[mintlify]

Preview deployment for your docs. Learn more about Mintlify Previews.

Project	Status	Preview	Updated (UTC)
unevenlabs	🟢 Ready	View Preview	May 7, 2026, 8:28 AM

💡 Tip: Enable Workflows to automatically generate PRs for you.

[github-actions]

Claude finished @pedromcunha's task in 1m 45s —— View job

---

Code Review: PR #309 — docs(drift): draft from internal source

• Read the PR diff
• Read full file context for features/deposit-addresses.mdx
• Cross-check against CLAUDE.md style rules
• Post review

---

This PR documents a new use of refundTo — passing the origin chain's "native-currency address" to auto-refund the depositor. The substance is a useful addition, but a few things should be tightened before merging.

1. Param-table cell is too long (CLAUDE.md §2.7)

features/deposit-addresses.mdx:26 — the new refundTo row in the Key Parameters table is now ~95 words and contains the full address triplet plus a rejection rule plus the existing omission rule plus a cross-link. §2.7 explicitly says: "Keep cell descriptions concise; move long explanations into prose below the table."

Suggest trimming the row to a short pointer and letting the Refund Behavior section (where the same triplet already lives, line 918) be the source of truth. Something like:

Address to send refunds if the transfer fails. Required for strict deposit addresses and for routes that include a destination-side swap. Pass the origin chain's native-currency address to auto-refund the depositor. Omitting on open addresses disables automatic refund. See Refund Behavior.

Fix this →

2. <Note> rewrite is hard to parse

features/deposit-addresses.mdx:923 — the new version inserts a clarifying clause between em-dashes, but the result reads as a triple-em-dash construction and breaks sentence flow:

do not use "refund to sender" — neither an explicit user address you don't control nor the native-currency-address auto-refund opt-in — the sender address will be the exchange's hot wallet, not the user's.

Two problems: (a) the appositive sits between an em-dash and another em-dash that originally introduced the explanation, which makes "the sender address will be…" feel disconnected from what it explains; (b) "an explicit user address you don't control" wasn't a refundTo option the table just defined — only the native-currency opt-in is the new "refund to sender" mechanism. The CEX-hot-wallet rationale applies to anything that resolves to the depositor.

Recommend splitting into two sentences, e.g.:

If users may send from a centralized exchange, do not refund to the depositor — neither via the native-currency-address opt-in nor a user-supplied address you don't control. The sender will be the exchange's hot wallet, not the user's. Use an app-controlled address instead so your support team can handle the last mile.

3. The address triplet is repeated three times

The same 0x000…0 / bc1qqq…mql8k8 / 1111…1111 list appears verbatim at lines 26, 918, and 930. Within a single page this isn't a CLAUDE.md §2.5 snippet-rule violation, but it's a maintenance hazard — if any address ever changes, three places must be edited in lockstep. Recommend keeping the canonical list in the Refund Behavior table row (line 918) and replacing the others with cross-references.

4. Please verify the Bitcoin "native-currency address"

bc1qqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqmql8k8 — the EVM zero address and the Solana system-program address are well-established conventions for "native asset", but I'm not aware of a comparable canonical "native BTC" address. Worth double-checking with the source PR / solver that this is the exact string the solver actually accepts (case-sensitive, checksum-valid bech32) before this ships, since a typo here is silently broken for integrators.

5. Pre-existing nit (not introduced by this PR)

The Key Parameters table at line 21 uses backticks (`useDepositAddress`) in the Property column rather than the **bold** convention CLAUDE.md §2.7 prescribes. Not blocking this PR — flagging for a future cleanup pass.

---

Summary

Substance is good; the new option is a useful addition. Three things to address before marking ready for review: (1) shorten the param-table cell, (2) rephrase the <Note>, (3) verify the Bitcoin address. Repetition (4) and the pre-existing table convention (5) are softer suggestions.
· Branch: docs-drift/src-c6bd60f6fe-d25c601

[github-actions]

Claude finished @pedromcunha's task in 1m 41s —— View job

---

Code Review: PR #309 — docs(drift): draft from internal source

• Read the PR diff
• Read full file context for features/deposit-addresses.mdx
• Cross-check against CLAUDE.md style rules
• Post review

---

Scope note

After the merge from main (commit 64a3c45), this PR's effective delta vs origin/main is now a single line — the <Note> callout at features/deposit-addresses.mdx:923. The other changes from the original drift commit (8d8343b) — the Key Parameters table row, the Refund Behavior table row, and the Recommended Setup bullet — have all landed on main independently. So the previous review's points 1, 3, 4 (param-table cell length, address-triplet repetition, Bitcoin-address verification) are out of scope for this PR; they apply to the file as it sits on main and belong in a separate cleanup pass if anyone wants to pursue them.

The only point still in scope is #2: the <Note> rewrite.

1. The <Note> rewrite is harder to parse than what's already on main

features/deposit-addresses.mdx:923 — this is the entirety of the PR. Side-by-side:

main (before):

If users may send from a centralized exchange, do not use "refund to sender" (an explicit user address you don't control, or the native-currency-address auto-refund opt-in) — the sender address will be the exchange's hot wallet, not the user's.

PR (after):

If users may send from a centralized exchange, do not use "refund to sender" — neither an explicit user address you don't control nor the native-currency-address auto-refund opt-in — the sender address will be the exchange's hot wallet, not the user's.

The new version replaces a parenthetical with an em-dash appositive, which puts the reader inside a — neither … nor … — construction and then asks them to pick up the main clause again with "the sender address will be…". The result is a one-sentence triple-dash structure where the third clause feels disconnected from the first. The parenthetical version on main reads more cleanly because the qualifier is visually subordinated by the parens, leaving a clean do not X — Y arc.

This is also the only edit in the PR, so as it stands the change makes the prose slightly worse, not better. Two options:

Option A — revert this line. If there's no semantic intent behind the punctuation swap, drop the change and close/repurpose the PR. (Recommended.)

Option B — split into two sentences. If the goal is to make the qualifier more explicit than parens allow:

If users may send from a centralized exchange, do not refund to the depositor — neither via the native-currency-address opt-in nor a user-supplied address you don't control. The sender will be the exchange's hot wallet, not the user's. Use an app-controlled address instead so your support team can handle the last mile.

That puts the qualifier on the same side of the em-dash as the directive, ends the first sentence cleanly, and lets the rationale start a fresh sentence. Still under the 15-line cap (CLAUDE.md §2.5) and within the 2–3 sentence paragraph budget (§2.10).

Fix this →

2. Style-rule pass on the touched line

For completeness, the touched line is otherwise compliant:

• §2.5 callout-length cap (15 lines): well under. ✓
• §2.10 paragraph length (2–3 sentences): two sentences. ✓
• §2.3 voice: "you"-addressed, active. ✓
• §2.9 emphasis: bold on the imperative, no italics. ✓

Summary

Given the merge from main, this PR's only remaining change is the <Note> rewrite at line 923, and that rewrite arguably degrades readability vs. what's already on main. Recommend either reverting the line (closing the PR as no-op) or splitting the qualifier into a separate sentence before marking ready for review.

· Branch: docs-drift/src-c6bd60f6fe-d25c601 ·