Problem
The fetcher script is trying to fetch markdown from URLs like:
https://platform.claude.com/docs/en/api/kotlin/messages.md
But platform.claude.com only serves HTML pages at URLs like:
https://platform.claude.com/docs/en/api/kotlin/messages
Impact
- All API documentation updates fail
- If accidentally merged, would replace 182+ API docs with HTML 404 pages
- Users would have completely broken API integration docs
Root Cause
scripts/fetcher.py:140 adds .md suffix to platform URLs:
fetch_url = f"{url}.md"
async with session.get(fetch_url) as response:
Solution Needed
The fetcher needs to distinguish between:
- code.claude.com - serves raw markdown at .md URLs ✅
- platform.claude.com - serves HTML pages, no .md URLs ❌
Either:
- Find the correct raw markdown API endpoint for platform docs
- Parse HTML content and convert to markdown
- Remove platform.claude.com docs from fetching until proper endpoint found
Closed PRs
💀 WHY URGENT: Would break API docs for all Claude Code users
Problem
The fetcher script is trying to fetch markdown from URLs like:
https://platform.claude.com/docs/en/api/kotlin/messages.mdBut platform.claude.com only serves HTML pages at URLs like:
https://platform.claude.com/docs/en/api/kotlin/messagesImpact
Root Cause
scripts/fetcher.py:140adds.mdsuffix to platform URLs:Solution Needed
The fetcher needs to distinguish between:
Either:
Closed PRs