Add /status endpoint with per-network health and metrics

[anhthii]

Summary

Add a GET /status HTTP endpoint that exposes per-network indexing health and metrics, identified by network_id. Designed for Grafana dashboards and operational monitoring.

Motivation

The current /health endpoint returns a simple {"status":"ok"} with no per-network visibility. Operators need to monitor individual chain health, detect lagging networks, and alert on degraded indexing performance.

Health Model

Health is derived per-network from total pending blocks (head gap + catchup backlog):

Status	Condition
healthy	pending_blocks < 50
slow	pending_blocks < 250
degraded	pending_blocks >= 250

Pending blocks calculation

head_gap = latest_block - indexed_block
catchup_pending_blocks = sum of remaining blocks across all active catchup ranges
pending_blocks = head_gap + catchup_pending_blocks

Data Source

Worker-cached — no additional RPC calls. The RegularWorker already fetches chain head and tracks currentBlock every tick. Workers push state into a shared in-memory Registry. The /status handler reads from that cache (at most one poll interval stale).

Response Shape

GET /status

{
  "timestamp": "2026-03-25T12:00:00Z",
  "version": "1.0.0",
  "networks": [
    {
      "network_id": "abc-123",
      "chain_name": "ethereum_sepolia",
      "internal_code": "ETHER_SEPOLIA_TESTNET",
      "network_type": "evm",
      "health": "healthy",
      "latest_block": 1000000,
      "indexed_block": 999998,
      "pending_blocks": 2,
      "head_gap": 2,
      "catchup_pending_blocks": 0,
      "catchup_ranges": 0,
      "failed_blocks": 0,
      "last_indexed_at": "2026-03-25T11:59:55Z"
    },
    {
      "network_id": "def-456",
      "chain_name": "tron_mainnet",
      "internal_code": "TRON_MAINNET",
      "network_type": "tron",
      "health": "degraded",
      "latest_block": 5000000,
      "indexed_block": 4999800,
      "pending_blocks": 10200,
      "head_gap": 200,
      "catchup_pending_blocks": 10000,
      "catchup_ranges": 1,
      "failed_blocks": 3,
      "last_indexed_at": "2026-03-25T11:59:30Z"
    }
  ]
}

Implementation Plan

File	Change
internal/status/status.go	NEW — types (HealthStatus, NetworkStatus, StatusResponse), Registry (thread-safe in-memory store), health derivation
internal/worker/manager.go	Add registry field + Registry() getter
internal/worker/factory.go	Create registry, register chain configs, pass to Manager
internal/worker/base.go	Add registry field to BaseWorker
internal/worker/regular.go	Call registry.Update() after processing blocks each tick
cmd/indexer/main.go	Register /status route, pass manager to health server

Grafana-Relevant Fields

All fields are scoped by network_id for dashboard filtering:

• pending_blocks — total unprocessed blocks (health signal)
• head_gap — real-time lag at chain head
• catchup_pending_blocks — historical backfill remaining
• catchup_ranges — number of active catchup ranges
• failed_blocks — blocks that failed processing
• indexed_block / latest_block — absolute block positions
• last_indexed_at — freshness of last successful index