fix(cli): honor -c cache_mode instead of always bypassing the cache (#2365)

[ntohidi]

Summary

Fixes #2365. crwl ignored -c cache_mode=... and always bypassed the cache, for two reasons:

1. --bypass-cache was a flag with default=True, so it always set CacheMode.BYPASS after -c and config files were applied.
2. -c and YAML/JSON configs pass cache_mode as a string ("enabled"). CacheMode is a plain Enum, so the string never matched and the cache was never read or written.

--bypass-cache now defaults to False. CrawlerRunConfig already defaults to BYPASS, so plain crwl still skips the cache. CrawlerRunConfig now converts a string cache_mode to CacheMode, and an invalid value raises an error.

crwl crawl https://example.com -o all -c cache_mode=enabled   # cache_status: miss
crwl crawl https://example.com -o all -c cache_mode=enabled   # cache_status: hit
crwl crawl https://example.com -o all                          # cache_status: miss (bypass)

List of files changed and why

• crawl4ai/cli.py: --bypass-cache defaults to False in crwl and crwl crawl.
• crawl4ai/async_configs.py: convert a string cache_mode to CacheMode.
• tests/cli/test_cli.py: default is still BYPASS; -c cache_mode=enabled gives ENABLED; -bc still forces BYPASS.

test_examples, test_missing_url and test_basic_crawl in tests/cli/test_cli.py already fail on develop; this PR does not change them.