feat: Implement Request ID Tracing for Tool Execution Chains (#107)

[TochukwuJustice]

closes #107
Summary of Changes:

Context Layer (src/logger.ts): Exported requestContext, a typed AsyncLocalStorage<{ requestId: string }>. Added a Pino mixin() that automatically injects request_id into every log entry emitted during a request — no additional code needed at any call site.
Tracing Middleware (src/index.ts): Wrapped each tool execution in requestContext.run(). A request_id is extracted from tool arguments if provided by the client; otherwise a cryptographically unique UUID is auto-generated via node:crypto.
Header Propagation (src/services/horizon.ts): Updated getHorizonServer() to read the active request context and forward X-Request-ID to every outgoing Stellar Horizon API call, enabling end-to-end tracing.
Unit Tests (tests/unit/tracing.test.ts): 4 tests covering context isolation, propagation, lifecycle, and concurrent request safety — all passing.
Reason for Changes: Complex multi-tool execution chains in MCP servers are difficult to debug without a consistent correlation ID. This implementation establishes a zero-overhead tracing layer that links logs, tool execution, and outgoing Stellar network requests under a single request_id.

[drips-wave]

@TochukwuJustice 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

[adams813]

Nice implementation, lgtm!

[TochukwuJustice]

Nice implementation, lgtm!

thanks man.. please merge PRs

[adams813]

Nice Implementation, LGTM!