Fetch and store full article text for content analysis

[alyssafuward]

Closes #69

Summary

• Adds a fetch-articles command that pulls the full article body (body_html) plus subtitle, tags, wordcount, and audience per post via /api/v1/posts/{slug}
• Stores an HTML-to-markdown conversion (body_markdown) alongside it in the posts table so the archive is directly queryable/readable
• Reuses existing SUBSTACK_SID auth and the project's hard-stop-on-429 rate limiting (no retries/backoff)
• Skips posts that already have a body unless --force is passed

Test plan

• python check.py — all 9 checks pass
• Ran python scraper.py fetch-articles end-to-end across both publications — 113/113 posts fetched, ~104K words total
• Confirmed rate-limit hard-stop triggers and resumes cleanly on rerun
• Confirmed no article content, cookies, or account info land in the diff (replies.db/config.py stay gitignored)

🤖 Generated with Claude Code