feat(reporter): add pluggable error reporter with noop default

[Abdulrasaq1515]

• Add lib/reporter.ts with reportError, configureReporter, resetReporter
• Wire SepError constructor to auto-call reportError with code + httpStatus
• Add tests covering noop default, mock reporter, Sentry-compatible pattern

Closes #184

Summary

Linked issue

Closes #184

Changes

Testing notes

Automated

• npm run typecheck · ⏳ not run / ✅ green / ❌ failing
• npm run lint · ⏳ not run / ✅ green / ❌ failing
• npm run test · ⏳ not run / ✅ green / ❌ failing
• npm run build · ⏳ not run / ✅ green / ❌ failing

New / modified tests

Manual verification (if applicable)

Screenshots / recordings

Before	After

Checklist

Correctness

• The PR title follows Conventional Commits (auto-linted)
• One logical change; unrelated cleanup was split into a separate PR
• npm run typecheck passes
• npm run lint passes with zero new warnings (we run --max-warnings 0 in CI)
• npm run test passes; new behaviour has a test
• npm run build passes

Data integrity

• No fabricated rates, stub prices, or placeholder exchange rates (see issue.md #005)
• No isMock, // MOCK, // TODO: replace with real data, or commented-out real code
• If touching an anchor: the anchor's stellar.toml is publicly resolvable at https://{domain}/.well-known/stellar.toml and contains TRANSFER_SERVER_SEP0024
• If touching SEP-10: network passphrase assertion is intact (mainnet only)
• If touching SEP-24: the 10s AbortController timeout is intact on anchor fetches
• If touching the status poll: terminal states (completed | refunded | error) still stop the SWR loop

Security & non-custody (see docs/NON_CUSTODY.md once it lands)

• No new code path holds user keys, user funds, or long-lived anchor JWTs
• Every signing action is performed by the user's wallet (Freighter today)
• No secrets committed; .env.local is unchanged; new env vars are added to .env.example

Docs

• User-facing behaviour change → CHANGELOG.md entry under [Unreleased]
• API / schema change → relevant docs/*.md updated in the same PR
• Architecture change → docs/ARCHITECTURE.md updated (file map, diagram, or invariants as applicable)
• Public-facing feature → screenshot added to docs/showcase/images/ when relevant
• New env var → .env.example + README env table updated

Release hygiene

• If this touches a wave deliverable, the matching [ ] in docs/ROADMAP.md is updated
• No dependency added without justification in the PR description
• No breaking change hidden inside a non-breaking commit

Breaking changes

None.

For reviewers

[vercel]

@Abdulrasaq1515 is attempting to deploy a commit to the ezedikeevan's projects Team on Vercel.

A member of the Team first needs to authorize it.

[ezedike-evan]

The implementation in lib/reporter.ts is clean and well-structured — the noop default, the configureReporter/reportError/resetReporter API, and the Sentry-compatible doc example are all good patterns. Tests are solid too.

Two things need fixing before this can merge:

1. No linked issue.
Closes #pluggable-error-reporter is a text string, not a GitHub issue number. Every PR in this project must close a tracked issue. If no issue exists yet for this feature, open one first, then reference it here with Closes #<number>. The PR template's "Linked issue" section is also left blank and the checklist is unchecked — please fill those in.

2. Auto-reporting in the SepError constructor is too aggressive.

// lib/stellar/errors.ts
reportError(this, { code, httpStatus })  // fires on every SepError construction

parseSepErrorBody is called whenever any anchor returns a non-2xx response — 404s, 400s, rate limits, validation errors. Wiring reportError unconditionally into the constructor means a Sentry-configured reporter would receive an alert for every routine anchor error. In production, a single user session hitting a throttled anchor could generate dozens of Sentry events.

The fix is straightforward: remove the auto-call from the constructor and instead call reportError at the specific call sites where the error is genuinely unexpected (e.g., a 5xx from the anchor, a timeout, a network failure). Let the caller decide what is worth reporting:

// In the caller that handles a real server error:
try {
  await submitChallenge(...)
} catch (err) {
  if (err instanceof SepError && err.httpStatus >= 500) {
    reportError(err, { anchorDomain })
  }
  throw err
}

The reporter module itself is good — just fix the issue link and the wiring point, and this is ready to merge.

[drips-wave]

@Abdulrasaq1515 Great news! 🎉 Based on an automated assessment of this PR, the linked Wave issue(s) no longer count against your application limits.

You can now already apply to more issues while waiting for a review of this PR. Keep up the great work! 🚀

Learn more about application limits