the llms.txt file from turborepo is literally 100k+ tokens
uhm guys I thought we standardized this
- the llms.txt file from turborepo is literally 100k+ tokensuhm guys I thought we standardized this
what are we doing? shouldn't this be split into multiple files and a general file that guides the llm to all other files? same thing as sitemaps - you have a root sitemap that links to other, more specific sitemapsi imagine doing this: llms/index.txt (contains structure) llms/introduction.txt llms/package-managers.txt llms/platforms.txt ...and so on
you're not using Context7 for accessing llms.txt with MCP?
