Harden CLI config parsing with strict validation and boundary tests

[Copilot]

Description

Daily audit requested identification and remediation of new bugs/security-quality risks. This PR addresses an untracked input-validation gap in CLI config parsing that allowed invalid runtime parameters to flow into training/simulation.

Changes in this pull request

• CLI validation hardening (src/main.py)

  • Rejects empty/whitespace-only --intervals
  • Enforces 0 < --test-size < 1
  • Enforces --target-days > 0, --n-trials > 0
  • Enforces --transaction-cost >= 0
  • Enforces --target-down-threshold < --target-up-threshold
  • Validates --target-horizons as a non-empty comma-separated list of positive integers
  • Normalizes interval parsing to drop empty comma entries

• Focused regression coverage (tests/test_main.py)

  • Adds parser tests for interval normalization and invalid-boundary rejection:
    • test-size = 0, 1, >1
    • target-days/n-trials = 0
    • negative transaction cost
    • inverted thresholds
    • zero/negative target horizons

if not 0 < args.test_size < 1:
    parser.error("--test-size must be between 0 and 1 (exclusive).")

if args.target_down_threshold >= args.target_up_threshold:
    parser.error("--target-down-threshold must be less than --target-up-threshold.")