Search in 📝MythOS is how creators, visitors, and AI agents find content across a knowledge library. The platform provides two complementary modes: keyword search for known-item lookup, and 📝semantic search for meaning-based retrieval.
How It Works
Keyword search matches memo titles and 📝hashtags. Type a word to find memos with that term in their title or tags. Prefix with # to restrict to exact tag matches — useful for browsing collections. Fast, precise, and available everywhere in the platform.
Semantic search goes deeper. It uses AI-generated vector embeddings to find memos by meaning, surfacing relevant content even when the exact words don't match. This powers the 📝MythOS MCP search_memos tool and the RAG pipeline behind 📝Chat with Library.
Key Capabilities
- Keyword search — matches titles and tags across the library. Prefix with # for exact tag matching
- Semantic search — AI-powered retrieval by meaning using vector embeddings and hybrid scoring. See 📝Semantic Search (MythOS)
- Tag collections — every hashtag creates a browseable collection. Search surfaces these collections alongside individual memos
- Permission-aware — results respect visibility settings. Private memos are only visible to the owner
- MCP integration — both search modes are available to AI agents via the
search_memosMCP tool
Getting Started
Keyword search is built into the MythOS interface — use the search bar in the sidebar. For semantic search via AI, connect your library via 📝MCP and see 📝How to Use Semantic Search in MythOS.
Related
- 📝Semantic Search (MythOS) — feature overview for AI-powered meaning-based search
- 📝Semantic Search API — technical documentation for the search interface
- 📝How to Use Semantic Search in MythOS — step-by-step tutorial for semantic search
- 📝How to Chat with Your Library via MCP — RAG-powered chat built on semantic retrieval
- 📝MythOS Features — full feature index
Thoughts
Search is the seam between having a library and using a library. Keyword search serves the organized mind — you know what you're looking for and you go get it. Semantic search serves the creative mind — you have a question and you trust the library to surface what's relevant. Both are necessary. Neither is sufficient alone.