webhook feature implemented

[feyishola]

closes #28

Summary of my implementation
src/webhook.rs

WebhookPayload — serializable event envelope with idempotency_key derived from tx hash + event index
DeadLetterEntry — captures failed deliveries (url, payload, attempts, last_error, failed_at)
WebhookDispatcher — contacts URLs sequentially in registration order, never skipping on failure
deliver_with_retry — exponential backoff: min(base * 2^attempt, max) + jitter(capped/4)
dispatch_background(Arc, event) — fire-and-forget via tokio::spawn
drain_dlq() / dlq_depth() — DLQ management, bounded at 10 000 entries
from_app_config — wires directly from AppConfig
src/metrics.rs

Added webhook_deliveries_total{status}, webhook_delivery_latency_seconds{status}, webhook_dlq_depth, webhook_retries_total
src/config.rs

5 new fields: webhook_max_retries (5), webhook_retry_base_delay_ms (200 ms), webhook_retry_max_delay_ms (30 s), webhook_request_timeout_ms (10 s), webhook_jitter_enabled (true)
Validation: base ≤ max delay, non-zero delays, boolean parsing
Corresponding env vars: WEBHOOK_MAX_RETRIES, WEBHOOK_RETRY_BASE_DELAY_MS, WEBHOOK_RETRY_MAX_DELAY_MS, WEBHOOK_REQUEST_TIMEOUT_MS, WEBHOOK_JITTER_ENABLED
src/lib.rs — registered pub mod webhook

src/main.rs — WebhookDispatcher in AppState, GET /webhooks/dlq, POST /webhooks/dlq/drain

README.md — added Webhook Delivery section with event payload schema, header table, retry formula, ordering/DLQ semantics, updated env-var table and metrics tables with alerting rules

[MaryammAli]

LGTM
thanks for ur contribution