improve logging and simplify deployment routing behavior

[jimmieme]

Summary

This PR improves space-agent observability and stability for simplified
deployments, especially LAN-first and compose-driven scenarios.

What Changed

• Added structured HTTP access logs (method, URI, status, latency, client IP,
  request ID).
• Replaced scattered fmt output with consistent logger-based messages.
• Added EnableDockerManage to allow running agent APIs without forcing docker
  service orchestration.
• Added URL normalization for internal service calls in single-docker mode
  (localhost -> compose service name).
• Improved /agent/info host persistence with host validation (skip localhost/
  loopback/empty hosts).
• Added fallback logic for BTID/SN/QR generation when hardware values are
  missing.
• Added router contract tests to ensure key internal/external APIs remain
  registered.

Why

• Simplified deployment mode needs predictable routing and better logs for
  troubleshooting.
• Existing behavior could bind to invalid loopback addresses in containerized
  scenarios.
• Missing hardware identifiers should degrade gracefully instead of breaking
  bind/QR flows.

Validation

• Verified agent starts with improved startup/access logs.
• Verified router contracts via routes_contract_test.go.
• Verified LAN host handling and fallback identity generation paths in
  simplified deployment flows.

Generated-by: Codex