fix: use canonical DER encoding for WebCrypto ECDSA signatures

[ajhodges]

Summary

Fixed a bug where WebCrypto ECDSA signatures were incorrectly DER-encoded, causing Go's strict encoding/asn1 parser to reject them.

The custom asn1EncodeSignature function didn't add the required 0x00 padding byte when an INTEGER's first byte has its high bit set (≥ 0x80). In DER encoding, this padding is mandatory to indicate the value is positive—without it, parsers interpret the value as negative.

Changes:

• Replaced custom asn1EncodeSignature with Ox's Signature.toDerBytes which produces canonical DER encoding
• Removed the buggy asn1EncodeSignature function and its unused imports
• Added comprehensive tests to verify correct DER encoding behavior

How did you test your changes?

Added unit tests in encoding.test.ts that verify:

• 0x00 padding is added when r has high bit set (e.g., 0x8e...)
• 0x00 padding is added when s has high bit set (e.g., 0xca...)
• No padding is added when high bit is not set (e.g., 0x6e..., 0x4a...)
• Small values with high bit set still get correct padding
• Edge cases with small r/s values work correctly

Tests fail with the old implementation (missing required padding) and pass with the fix.

Also ran playground locally and made a few cryptokey transactions via http://localhost:3001/add-sub-account, which seemed to work fine

[cb-heimdall]

✅ Heimdall Review Status

Requirement	Status	More Info
Reviews	✅ 2/1	Denominator calculation Show calculation 1 if user is bot 0 1 if user is external 0 2 if repo is sensitive 0 From .codeflow.yml 1 Additional review requirements Show calculation Max 0 0 From CODEOWNERS 0 Global minimum 0 Max 1 1 1 if commit is unverified 0 Sum 1

[montycheese]

verified works locally, thanks for finding this!

[montycheese]

nice!

[stephancill]

neat, thanks!