Wiki and exports
Export the graph as an agent-crawlable markdown wiki, an Obsidian vault, an interactive HTML viewer, or GraphML/JSON/Neo4j.
Semantic enrichment
Two optional semantic layers — local embeddings (no API key) and LLM enrichment — that add similar_to edges and concept nodes to the graph.
Agent integration
Connect AI coding assistants to the graph — MCP server, skill-file installers for ten platforms, git hooks, and the editor-side hook-guard.
Global graph (cross-repo)
Merge many repo graphs into one queryable store at ~/.astria/global.db — shared symbols deduplicate, cross-repo calls resolve.
Memory and learning
Two feedback loops — automatic learned edges from query history, and curated Q/A memory via save-result and reflect.