How to find memos by meaning — not just keywords — in your 📝MythOS library using 📝semantic search.

When your library has hundreds of memos, keyword search only finds what you remember naming. Semantic search finds what you forgot you wrote — it understands the meaning behind your query and matches it against the actual content of your memos.

What You'll Need

• A MythOS library connected to an AI client via 📝MCP (see setup tutorials for 📝Claude.ai, 📝Claude Code, or 📝Cursor / Claude Desktop)

Steps

1. Ask your AI — to search your library semantically. Example: "Search my MythOS library for memos about distributed systems." Your AI calls search_memos with mode: "semantic" behind the scenes
1. Use rich queries — describe what you're looking for in natural language. "How does authentication work in my app" performs better than "auth." Semantic search matches meaning, so more context gives better results
1. Review the results — each result includes the memo title, a relevance score, and the specific text chunks that matched your query. Higher scores mean stronger semantic matches
1. Refine if needed — narrow results by adding a tags filter. Example: "Search semantically for deployment, filtered to the mythos tag." Or increase the limit to surface more candidates
1. Follow the trail — click through to the memos that match. The chunks shown in results tell you exactly which part of the memo is relevant, so you can jump straight to the useful section

What's Next?

• 📝How to Chat with Your Library via MCP — ask questions answered by your library's content via RAG
• 📝Search (MythOS) — overview of all search capabilities including keyword search
• 📝Semantic Search API — technical documentation for the search interface
• 📝MythOS Content Topology — structure your library for better retrieval

Contexts

• 🏷️#mythos
• 🏷️#mythos-feature