fix: use runtime factory for eval entrypoint discovery and show usage help

[cristipufu]

Summary

• Replace hardcoded auto_discover_entrypoint() (uipath.json-only) with runtime_factory.discover_entrypoints() which supports all runtimes (langgraph, llama_index, etc.)
• When auto-discovery fails (multiple entrypoints or eval sets), show a clean usage message listing available options with example commands instead of a ValueError traceback
• Remove unused auto_discover_entrypoint and auto_discover_eval_set functions

Test plan

• Added 7 new tests in tests/cli/eval/test_eval_discovery.py
• Multiple entrypoints shows usage help with all available entrypoints and eval sets
• Multiple eval sets shows usage help with full relative paths
• Single entrypoint + eval set auto-discovers without showing help
• No entrypoints shows helpful guidance message
• Explicit entrypoint still triggers eval set discovery
• All 1514 existing tests pass

🤖 Generated with Claude Code

Development Package

• Use uipath pack --nolock to get the latest dev build from this PR (requires version range).
• Add this package as a dependency in your pyproject.toml:

[project]
dependencies = [
  # Exact version:
  "uipath==2.10.4.dev1013985129",

  # Any version from PR
  "uipath>=2.10.4.dev1013980000,<2.10.4.dev1013990000"
]

[[tool.uv.index]]
name = "testpypi"
url = "https://test.pypi.org/simple/"
publish-url = "https://test.pypi.org/legacy/"
explicit = true

[tool.uv.sources]
uipath = { index = "testpypi" }

[tool.uv]
override-dependencies = [
    "uipath>=2.10.4.dev1013980000,<2.10.4.dev1013990000",
]