🧠 Engram - Semantic Search Skill

This skill enables an AI agent to perform semantic searches on a local folder of Markdown files (e.g., an Obsidian vault). It finds information based on the meaning and context of a query, not just exact keywords.

Tools

engram_search

Searches the indexed knowledge base.

• query (string, required): The natural language question to ask.
• top_k (number, optional): The number of results to return.
• min_score (number, optional): The minimum relevance score (0.0 to 1.0) for results.

engram_index

Builds or updates the search index from the local Markdown files. This tool should be run periodically to keep the search memory synchronized.

Author

• Andrie Wijaya (@Anwitch)