Document scrape freshness and cache metadata

[hmishra2250]

Summary

• document existing firecrawl scrape --max-age and --json in the scrape skill
• add fresh-retrieval/cache provenance guidance for final verification scrapes
• add job/ATS liveness guidance for agent workflows

Root cause / behavior change

CLI agents could already pass --max-age, but the skill did not teach when to request a fresh retrieval path or how to inspect cache provenance. The scrape skill now makes the final-verification pattern explicit without treating retrieval freshness as source-object liveness.

Validation

• scoped grep for --max-age, --json, metadata.cache, ATS/liveness guidance ✅
• git diff --check ✅
• GitHub Actions ✅: tests, macOS/Linux/Windows binary tests, and macOS/Linux/Windows npm package smoke tests