Labels: Stellar Wave, stellar, test, crypto, drips, help-wanted, good-first-issue
Tier: M (2–4 days)
Type: test
Context
Stealth address math involves a lot of scalar arithmetic mod L (the ed25519 group order). Unit tests cover specific values, but property-based tests are far better at catching the edge cases that break in production — overflow, wrap-around, near-boundary inputs, zero/identity, etc.
This issue adds fast-check-driven property tests for sdk/src/chains/stellar/scalar.ts.
Scope
Add fast-check as a dev-dependency and write property tests under test/chains/stellar/properties.test.ts:
Properties to cover
For all valid scalars a, b, c and bytes x in the appropriate ranges:
- Addition associativity:
(a + b) + c == a + (b + c) mod L
- Addition commutativity:
a + b == b + a mod L
- Additive identity:
a + 0 == a mod L
- Reduction stability:
bytesToScalar(scalarToBytes(a)) == a for all valid a
seedToScalar determinism: same seed → same scalar; different seeds → different scalars (with negligible collision probability)- Stealth scalar correctness: for random
(m, s_h), (m + s_h) * G == m*G + s_h*G (the equation that makes stealth spending work)
- View-tag uniformity: across 10k random shared secrets, the view tag distribution chi-square test passes against uniform[0,255]
signWithScalar round-trip: for any scalar a and message m, verify(signWithScalar(a, m), m, a*G) == true
Configuration
- Use
fast-check's BigInt-aware arbitraries for the scalar space.
- Default test count: 1000 cases. Add a
pnpm test:fuzz script that bumps to 100,000.
- Add a CI job (slow-tests) that runs the high-case version nightly.
Acceptance criteria
Why this is a good first issue
The math is well-defined, the harness is contained to one file, and the acceptance criteria are objective (tests pass or don't). Good ramp for someone learning the Wraith codebase before tackling the larger audits.
Resources
Labels:
Stellar Wave,stellar,test,crypto,drips,help-wanted,good-first-issueTier: M (2–4 days)
Type: test
Context
Stealth address math involves a lot of scalar arithmetic mod L (the ed25519 group order). Unit tests cover specific values, but property-based tests are far better at catching the edge cases that break in production — overflow, wrap-around, near-boundary inputs, zero/identity, etc.
This issue adds
fast-check-driven property tests forsdk/src/chains/stellar/scalar.ts.Scope
Add
fast-checkas a dev-dependency and write property tests undertest/chains/stellar/properties.test.ts:Properties to cover
For all valid scalars
a, b, cand bytesxin the appropriate ranges:(a + b) + c == a + (b + c)mod La + b == b + amod La + 0 == amod LbytesToScalar(scalarToBytes(a)) == afor all validaseedToScalardeterminism: same seed → same scalar; different seeds → different scalars (with negligible collision probability)(m, s_h),(m + s_h) * G == m*G + s_h*G(the equation that makes stealth spending work)signWithScalarround-trip: for any scalaraand messagem,verify(signWithScalar(a, m), m, a*G) == trueConfiguration
fast-check'sBigInt-aware arbitraries for the scalar space.pnpm test:fuzzscript that bumps to 100,000.Acceptance criteria
Why this is a good first issue
The math is well-defined, the harness is contained to one file, and the acceptance criteria are objective (tests pass or don't). Good ramp for someone learning the Wraith codebase before tackling the larger audits.
Resources
fast-checkdocs: https://fast-check.dev/test/chains/stellar/scalar.test.ts