from the export types script

* feat: Support for PQ accounts (WIP)

This work was originally made with claude assistance and adapted from there.

* chore: Remove testing env override

* chore: tidy go.sum for the patched sandbox and add cgo support (for falcon)

* chore: Use the master branch for the test-harness

* chore: Move falcon signatures under a new build constraint

* chore: Use the falcon buildtag for go vet

* feat: Add an extensible signer API

* chore: Sloppy claude tests for the old api surface

The real logic is already covered by the old tests we migrated

* feat: Add an extensible signer API for falcon

* chore: Remove internalFalcon1024Signer (use SaltedFalcon1024Signer instead)

* chore: Add a note for VerifyLogicSig

* chore: There's no reason for Falcon1024Salt() to fail

* chore: Fix a bunch of typos

* chore: Small fixups

* chore: Add ed25519 and delegation operations to transactionSigner.go

* chore: Move everything PQ to the PQ prefix (and use Falcon1024 only for explicit falcon stuff)

* chore: go vet should work both with and without the falcon tag

* chore: Rename Falcon1024AccountTransactionSigner to PQAccountTransactionSigner

* fix: Wrong code on PQAccountTransactionSigner#Equals

* chore: Improve VerifyLogicSig docs

* chore: Improve README note on the `falcon` tag

* refactor: move transaction signing wrappers out of crypto

Move scheme-specific signed transaction construction into the
transaction package and expose a scheme-neutral SignTransaction helper.
Preserve the legacy signing APIs and public API compatibility.

* chore: various fixes from review and empty signer

* refactor: deduplicate signing helpers across crypto and transaction

Collapse the repeated signature-length checks, delegation-signature counting, JSON-based signer equality, and per-index signing loops into shared helpers within each package.

Two of the deduplicated call sites, the program signers in makeLogicSig and Ed25519AppendMultisigToLogicSig, previously copied the raw signature without checking its length, silently accepting a short signature from a custom Ed25519Signer as a subsig. They now return errInvalidSignatureReturned like every other signing path.

Move the PQ golden tests to the transaction package so they exercise PQAccountTransactionSigner directly, and drop signPQAccountTransactionFixture, a test-only reimplementation of the signing envelope that could have passed while the real signer was broken. Add coverage for the rekeyed PQ signer.

* chore: review feedback

* chore: further review feedback

* chore: code review

* test: fix var reassignment and comments

* chore: remove non-canonical salt support

* feat: use pointers for signer equality check

This more closely matches the behavior of JS and Py. The idea is that
one address could have multiple signers, although I'm not sure how
common that actually is. This definitely feels more correct, but if
there are downsides I'm not seeing we could revert it

* feat: falcon seed size check

While this technically isn't required, it prevents keys from being
generated with not enough entropy

* feat: hasPQSig in SignatureCount

This is technically a breaking change, but precedent was set when adding
LMSig. Presumably this is not commonly used by external users

* refactor: internal/signing

* chore: rename SaltForPQSigner to CanonicalSaltForPQAddress

* chore: use conditional build tags

* chore: remove nil signer err

* chore: remove nil sgnr check

* chore: simplify pq empty signer

* nit: fix readme

* nit: do auth addr on a single place

* chore: Don't require a full signer when using PQEmptyTransactionSigner

---------

Co-authored-by: Joe Polny <joepolny@gmail.com>

chore(deps): bump cloudflare/wrangler-action from 4.0.0 to 4.1.3 in the minor-and-patch group